Document 10534746

advertisement

AN ABSTRACT OF THE DISSERTATION OF

Tara M. Hudiburg for the degree of Doctor of Philosophy in Forest Science presented on

June 14, 2012.

Title: Analysis of the Regional Carbon Balance of Pacific Northwest Forests Under

Changing Climate, Disturbance, and Management for Bioenergy

Abstract approved:

Beverly E. Law

Atmospheric carbon dioxide levels have been steadily increasing from anthropogenic energy production, development and use. Carbon cycling in the terrestrial biosphere, particularly forest ecosystems, has an important role in regulating atmospheric concentrations of carbon dioxide. US West coast forest management policies are being developed to implement forest bioenergy production while reducing risk of catastrophic wildfire. Modeling and understanding the response of terrestrial ecosystems to changing environmental conditions associated with energy production and use are primary goals of global change science. Coupled carbon-nitrogen ecosystem process models identify and predict important factors that govern long term changes in terrestrial carbon stores or net ecosystem production (NEP). By quantifying and reducing uncertainty in model estimates using existing datasets, this research provides a solid scientific foundation for evaluating carbon dynamics under conditions of future climate change and land management practices at local and regional scales. Through the combined use of field observations, remote sensing data products, and the NCAR CESM/CLM4-CN coupled carbon-climate model, the objectives of this project were to 1) determine the interactive effects of changing environmental factors (i.e. increased CO2, nitrogen deposition, warming) on net carbon uptake in temperate forest ecosystems and 2) predict the net carbon emissions of West Coast forests under future climate scenarios and implementation of bioenergy programs. West Coast forests were found to be a current strong carbon sink after accounting for removals from harvest and fire. Net biome production (NBP) was 26 ± 3 Tg C yr

-1

, an amount equal to 18% of Washington, Oregon, and California fossil fuel emissions combined. Modeling of future conditions showed

increased net primary production (NPP) because of climate and CO

2 fertilization, but was eventually limited by nitrogen availability, while heterotrophic respiration (R h

) continued to increase, leading to little change in net ecosystem production (NEP). After accounting for harvest removals, management strategies which increased harvest compared to business-as-usual (BAU) resulted in decreased NBP. Increased harvest activity for bioenergy did not reduce short- or long-term emissions to the atmosphere regardless of the treatment intensity or product use. By the end of the 21 st

century, the carbon accumulated in forest regrowth and wood product sinks combined with avoided emissions from fossil fuels and fire were insufficient to offset the carbon lost from harvest removals, decomposition of wood products, associated harvest/transport/manufacturing emissions, and bioenergy combustion emissions. The only scenario that reduced carbon emissions compared to BAU over the 90 year period was a ‘No Harvest’ scenario where NBP was significantly higher than BAU for most of the simulation period. Current and future changes to baseline conditions that weaken the forest carbon sink may result in no change to emissions in some forest types.

©Copyright by Tara M. Hudiburg

June 14, 2012

All Rights Reserved

Analysis of the Regional Carbon Balance of Pacific Northwest Forests Under Changing

Climate, Disturbance, and Management for Bioenergy by

Tara W. Hudiburg

A DISSERTATION submitted to

Oregon State University in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Presented June 14, 2012

Commencement June 2013

Doctor of Philosophy dissertation of Tara W. Hudiburg presented on June 14, 2012.

APPROVED:

Major Professor, representing Forest Science

_____________________________________________________________________

Head of the Department of Forest Ecosystems and Society

Dean of the Graduate School

I understand that my dissertation will become part of the permanent collection of Oregon

State University libraries. My signature below authorizes release of my dissertation to any reader upon request.

Tara W. Hudiburg, Author

ACKNOWLEDGEMENTS

First of all, I would like to thank my major professor, Beverly Law, for supporting me and providing the necessary guidance in order for me to complete my degree. Bev’s faith in my ability and her encouragement in my pursuit of a PhD have been essential to my success. I would also like to thank Peter Thornton for patiently teaching me how to become an ecosystem ‘modeler’. I would also like to thank my other committee members, Richard Waring and Dominique Bachelet for their continued support and genuine interest in my research and success. I would like to extend a very special thanks to my peers at OSU and the greater scientific community, especially Sebastiaan Luyssaert for his quantitative assistance and mentorship and Terralyn Vandetta without whom I would never have been able to run a single model simulation on the Linux cluster.

Finally, I want to thank my family; my parents, Ron and JoAnne Hudiburg and my husband Kris. Your confidence, support, and encouragement were essential. And last of all, my children Quinlan and Mackenzie, for understanding and patience when I was unable to play with you because I was too busy working on my dissertation.

This research was supported by the U.S. Department of Energy Biological and

Environmental Research Terrestrial Carbon Program (Award # DE-FG02-04ER64361) and the DOE Global Change Graduate Fellowship. This study is part of a regional North

American Carbon Program project. Special thanks to Catharine van Ingen for her assistance with database organization and implementation, and to Microsoft for database design funding through an E-Science grant and a Microsoft Research graduate student internship. We would like to thank the PNW-FIA regional office for their assistance and cooperation in compiling the database and James Sulzman for programming consultation.

CONTRIBUTION OF AUTHORS

For Chapter 2, Tara Hudiburg designed and implemented the study with guidance from

Beverly Law and Sebastiaan Luyssaert. Tara Hudiburg, Sebastiaan Luyssaert, and

Beverly Law co-wrote the paper and Sebastiaan contributed to parts of the analysis.

Christian Wirth provided essential data and methods for the analysis and valuable comments on the manuscript. For Chapter 3, Tara Hudiburg and Beverly Law designed, implemented and co-wrote the paper. Tara Hudiburg did the analysis and Peter Thornton provided guidance and valuable comments on the manuscript. For Chapter 4, Tara

Hudiburg designed and implemented the study with guidance from Beverly Law and

Peter Thornton. Tara Hudiburg, Sebastiaan Luyssaert, Beverly Law, and Peter Thornton co-wrote the paper.

TABLE OF CONTENTS

Page

Chapter 1: Introduction………………………………………………………

Chapter 2: Regional CO

2

implications of forest bioenergy production …….

2

9

Chapter 3: Evaluation and improvement of the Community Land Model

(CLM 4.0) in Pacific Northwest Forests …………………………………..

Chapter 4: Long-term effects of bioenergy harvest on the carbon balance of

25

Pacific Northwest forests under changing climate and climate-related disturbance 57

Chapter 5: Conclusion……………………………………………………… 104

110 Bibliography………………………………………………………………

Appendices……………………………………………………………….

Appendix A Supporting Information for Chapters 2 and 4……..

122

123

160 Appendix B Supporting Information for Chapters 3 and 4……..

LIST OF FIGURES

Figure

2.1 Current net biome productivity (NBP), NBP uncertainty, and

Page change in NBP required for zero net emissions compared to business-as-usual (BAU)………………………………………….. 22

2.2. Life Cycle Assessment carbon emission trends by ecoregion for each scenario.………………………………………………….

2.3. Total regional carbon sinks, sources, and added emissions for

23 each scenario………………………………………………………. 24

3.1. Study region area (Oregon) divided by ecoregions………………. 51

3.2. Maps of modeled estimates of a) Stem Biomass, b) NPP, c) NEP, and d) heterotrophic respiration (Rh). …………………………….

3.3. Ecoregion means of modeled and observed evergreen needleleaf

(ENF) PFT values for the period from 2001-2006.……………….

52

53

3.4. Simulated monthly GPP versus observed GPP at two FLUNET eddy-covariance tower sites in the study region..…………………… 54

3.5. 3D surface plots of shortwave radiation versus GPP (LUE) versus monthly average temperature…………………………………………

55

3.6. Observed values for ENF leaf carbon to nitrogen ratios (A) and specific leaf area (B) from over 100 supplemental plots in the region.. 56

4.1. Study region area (Oregon, USA) divided by ecoregions and coded according to potential management plans. ……………………………. 89

4.2. Predicted regional change in: (A) annual temperature and precipitation from 2010 to 2100, (B) NPP (dashed lines) and R h

(solid lines), (C) NEP, and (D) mineralized soil nitrogen for each of the factors tested (climate,

CO

2

and N deposition, BAU harvest)…………………………….. 90

4.3 The change (delta) for each management scenario compared to BAU

in the state of Oregon for a) NPP, b) Rh, c) NEP, d) NBP, e) Total

Carbon, and f) Harvest………………………………………………… 91

Figure

LIST OF FIGURES (Continued)

Page

4.4. Oregon total net ecosystem production (NEP), fire emissions, wood harvest removals, and net biome production (NBP) in Tg C per year……. 92

4.5. The change (delta) for each management scenario compared to BAU

in the Blue Mountains for a) NPP, b) Rh, c) NEP, d) NBP, e) Total

Carbon, and f) Harvest…………………………………………………. 93

4.6. The change (delta) for each management scenario compared to BAU

in the Coast Range for a) NPP, b) Rh, c) NEP, d)NBP, e) Total Carbon, and f) Harvest……………………………………...……………….

94

4.7. The change (delta) for each management scenario compared to BAU

in the East Cascades for a) NPP, b) Rh, c) NEP, d)NBP, e) Total Carbon, and f) Harvest……………………………………...………………. 95

4.8. The change (delta) for each management scenario compared to BAU

in the Klamath Mountains for a) NPP, b) Rh, c) NEP, d)NBP, e) Total

Carbon, and f) Harvest………………………………...………………. 96

4.9. The change (delta) for each management scenario compared to BAU

in the West Cascades for a) NPP, b) Rh, c) NEP, d)NBP, e) Total Carbon, and f) Harvest……………………………………...………………. 97

4.10. Total change in Net C e

to the atmosphere summed over the 90 year period compared to BAU for each management scenario in each ecoregion in Tg C...…………………………………………………………… 98

4.11. Blue Mountains LCA by component for a) the Thinning Scenario only and b) All scenarios Net C e

compared to BAU (Delta BAU)..………… 99

4.12. Coast Range LCA by component for a) the Thinning Scenario only and b) All scenarios Net C e

compared to BAU (Delta BAU)..………… 100

4.13. East Cascades LCA by component for a) the Thinning Scenario only and b) All scenarios Net C e

compared to BAU (Delta BAU)..………… 101

4.14. Klamath Mountains LCA by component for a) the Thinning Scenario only and b) All scenarios Net C e

compared to BAU (Delta BAU)..………… 102

LIST OF FIGURES (Continued)

Figure Page

4.15. West Cascades LCA by component for a) the Thinning Scenario only and b) All scenarios Net C e

compared to BAU (Delta BAU)..………… 103

LIST OF TABLES

Table Page

3.1. Ecoregion and tower site characteristics in study region……………… 48

3.2. Historical model simulations………………………………………….. 49

3.3. Stem carbon and NPP reduced statistic and model bias by ecoregion, annual precipitation regime a

, and age group b

..................... 50

4.1. Ecoregion forested hectares percent of total study region forested

area, mean stand ages by ownership, dominant forest types, mean annual precipitation

(MAP), mean annual temperature (MAT), and proposed bioenergy management scenarios.……………………………….… 85

4.2. CLM4 future simulations for control, baseline, and bioenergy management scenarios…………………………………………….. 86

4.3 Ecoregion forested hectares percent of total study region forested area, mean stand ages by ownership, dominant forest types, mean annual precipitation (MAP), mean annual temperature (MAT), and proposed bioenergy management scenarios…………………………………….. 87

4.4. Business-as-Usual (BAU) regional carbon stocks and fluxes and soil mineralized nitrogen (Sminn) summed over the treatment period

(90 years) and the difference compared to BAU for each of the management scenarios........................................................................ 88

LIST OF APPENDIX FIGURES

Figure Page

A.1. Thinning treatment scenarios…………………………….………. 157

A.2. (A) MFRI (Mean Fire Return Interval) verses delta CO

2

(Fire Prevention;

FP – Business as Usual; BAU) and (B) initial emissions versus delta

CO

2

……………………………………………………………….

158

A.3. Conceptual analysis of harvest to NEP ratio, product mix, and varying

159 efficiencies of bioenergy production……………………………..

B.1. Stemwood allocation (ratio of stem wood allocation to foliage allocation) versus NPP.…………………………….…………… 166

B.2. GPP versus monthly precipitation at the flux towers sites. Red squares are tower observations of GPP and blue diamonds are the

CLM simulated monthly GPP values. …………………………….

B.3. Historical annual area burned in Oregon (km

-2

). Estimates are from the Monitoring Trends in Burn Severity database (MTBS; Eidenshink,

167

J. et al. 2007), the Global Fire Emissions Database (GFED; van der Werf,

G. et al. 2010), and CLM4. CLM4 overestimates burn area in Oregon for all years except for 2002 compared to the remote sensing based estimates.………………………………………………………….. 168

B.4. Simulated burn area by CLM4 for the RCP4.5 future climate scenario. 169

B.5. Predicted annual relative humidity (%) using the downscaled regional climate data and DAYMET algorithms. ……….……………………. 170

B.6. Monthly GPP and monthly precipitation for a period of drier years

(2076-2078) followed by a period of wetter years (2079-2081). ………….. 171

B.7. Monthly GPP and monthly precipitation for a period of drier years

(2076-2078) followed by a period of wetter years (2079-2081). ………….. 172

LIST OF APPENDIX TABLES

Table Page

A.1. Ecoregion characteristics including dominant forest types, mean annual temperature (MAT), mean annual precipitation (MAP), and area weighted mean fire return interval (MFRI).……………………………………. 144

A.2. State total and mean carbon fluxes for Business-As-Usual (BAU) and treatments………………………………………………………….

145

A.3. Forest types that contribute more than 1 g C m

-2

y

-1

to the observed area-weighted difference of 46.4 g C m

-2

y

-1

(see also Appendix A

Table 2) in net biome production (NBP) between California and Oregon for shared forest types…………………………………………………. 146

A.4. Life Cycle Assessment coefficients………………………….……. 148

A.5. Life cycle assessment of forest derived bioenergy for the West Coast region assuming a 50:50 mix of combustion for combined heat and power

(CHP) and conversion to cellulosic ethanol (CE). ……………….. 150

A.6. Sensitivity Analysis parameters and relative impact on results (CHP = biomass used as combined heat and power; FP = Fire Prevention scenario;

EC = Economically Feasible scenario; BP = Bioenergy Production scenario).…………………………………..……………………… 152

A.7. Forest Inventory Analysis plot information for the study region….. 153

A.8. Equations and factors used to calculate carbon stocks, fluxes, and life-cycle assessment………………………………………………… 154

A.9. Comparison of combustion factors by source and fuel category… 156

B.1. CLM PFT original configuration and the new subgroup that replaced the original PFT for regional variation in CLM_eco and CLM_stem….

163

B.2. Model physiological parameters that vary for subgroups of PFTs in each ecoregion………………………………………….……………. 164

B.3. Stem wood allocation equations for each ecoregion PFT……………...165

Analysis of the Regional Carbon Balance of Pacific Northwest Forests Under Changing

Climate, Disturbance, and Management for Bioenergy by

Tara W. Hudiburg

Chapter 1: Introduction

It is general knowledge that carbon dioxide (CO

2

), a greenhouses gas (GHG), has increased from ~260 to near 400 ppm in the Earth’s atmosphere since the Industrial

Revolution began in the early 1800’s as a result of the combustion of fossil fuels for

2 energy and conversion of native vegetation to other uses, primarily agriculture

( http://www.esrl.noaa.gov/gmd/ccgg/trends/history.html

). Incoming solar radiation that is not absorbed at the Earth’s surface is reflected back to the atmosphere as long wave radiation where it can be absorbed by GHGs such as CO

2

and water vapor. This reduces the amount of energy emitted to space and warms the planet. Because CO

2

absorbs energy at some wavelengths that water vapor misses, the extra absorption caused by increased concentrations of CO

2

warms the air more. The residence time of CO

2

in the atmosphere is centuries long so its effect on climate will continue for a millennium, even if emissions are immediately reduced back to pre-industrial levels (Solomon et al. 2009).

Plants absorb CO

2

in the process of photosynthesis, store it temporarily in their biomass, and release it slowly back to the atmosphere as organic matter decays. Forests store the most carbon (50% of dry mass) of any life form, largely because of the annual accumulation of tree wood. Currently, temperate forests in the United States capture

~20% of our fossil fuel emissions (Birdsey 2007). Whether the capacity of forests to sequester carbon can be maintained is a question faced by policy makers and land managers alike, particularly in the face of a changing climate. Disturbances by fire, insects, and diseases that can reduce forest carbon storage have significantly increased in recent years, and this trend is projected to continue (Rogers et al. 2011; van Mantgem et al. 2009; Kurz et al. 2008).

In efforts to improve forest health, create jobs, restore habitat, provide an alternative to fossil fuels, and mitigate climate change, management plans have been implemented to increase utilization of forest biomass for wood products and bioenergy by thinning, rotation-based harvests, and salvage logging. Thinning can be an effective strategy for preventing wildfire where risk is high, especially around homes, farms and industrial sites (Raymond and Peterson 2005, Stephens et al. 2009). Rotation-based

3 harvests provide a steady supply of wood for products and bioenergy and salvage logging makes use of wood that is no longer taking up carbon. The extent that these practices reduce the rate that forests sequester carbon is not well established but could be significant, particularly in regions such as the Pacific Northwest where the potential for biomass accumulation west of the crest of the Cascade Mountains is the highest recorded

(Waring and Franklin 1979). The Pacific Northwest, however, contains a wide range of environments representing almost the full range of productivity in North America. Any analysis of carbon dynamics must account for this spatial variation.

Net primary production (NPP) ranges from 100 to 900 g C m

-2

across the region and (Turner et al. 2007) net ecosystem production (NEP; the balance between photosynthesis and respiration) varies from 0 to 300 g C m

-2

yr

-1

(Williams et al. 2012).

After accounting for wood removals through fire and harvest (NBP; net biome production), Hayes et al., (2012) estimated that US west coast forests in recent years are a net carbon sink of between 10-20 Tg C yr

-1

, an amount near the current fossil fuel emissions of Oregon (Gurney et al. 2009).

Forest carbon sinks will change as climate and societal demands change. Harvest rates vary with wood product demand and conservation goals. Fire severity and burn area are increasing with warming climate (Westerling et al. 2006). There are more widespread mortality events in the western US states due to insect outbreaks (Raffa et al. 2008) and drought (Breshears et al. 2008, van Mantgem et al. 2009, McDowell et al. 2010). In order for policy makers to provide effective carbon sequestration strategies involving forests, current and potential forest carbon sinks should be quantified using a variety of climate and management scenarios.

There are a number of approaches available to evaluate the relative merits of various policies on carbon sequestration in forests and in their manufactured products.

Approaches range from small case studies of thinning combined with bioenergy to reduce fire and fossil fuel emissions (Winford and Gaither Jr. In Press), to national evaluation of wood product sinks (Lippke et al. 2010), and to continental assessments of combined forest and wood product carbon sinks (Hayes et al. 2012). Although the methods for quantifying net forest carbon uptake varies (i.e. observations, mathematical models,

simulation models, remote sensing, or combinations), the method for quantifying the amount of carbon eventually emitted to the atmosphere from wood product use is accomplished using life cycle assessment (LCA).

LCA is used to track forest carbon stocks and emissions in relation to bioenergy

4 production and wood products, however the number of variables used and accounted for differs widely. Some studies include emissions associated with product use and decomposition, but neglect the emissions generated for harvest, transportation, and manufacturing (Routa et al. 2011).Others use a complete LCA for the wood once it leaves the forest, but ignore the baseline carbon uptake and loss from the forest itself (in other words net uptake is assumed to be zero; Lippke et al. 2011). Clearly, removal of any wood reduces in-situ carbon storage. However, because of the inconsistencies concerning baseline conditions and LCA variables, there is continued debate about whether forest management strategies that include bioenergy will be carbon neutral (Lippke et al. 2011,

Malmsheimer et al. 2011), or will result in an increase in emissions compared to current management practices over the next several decades(Mitchell et al. 2012). In spite of this contention, bioenergy plans and facilities are continuing in the Western US because of asserted local economic considerations (Siemers 2011).

Management policies need to be designed to: (1) not further degrade the forest carbon sink, (2) target areas where it may be increased, (3) avoid detrimental increases in atmospheric CO

2

emissions, and (4) provide energy. The consequences of each management strategy must be compared to baseline conditions that include business-asusual (BAU) management practices. The overall goal of this study was to evaluate the short and long term effects of climate, disturbance, and potential bioenergy management strategies on regional carbon storage and emissions using existing datasets and an internationally recognized terrestrial ecosystem process model (CLM4; Lawrence et al.

2011).

Chapter 2 addresses the short-term effects of bioenergy management on net CO

2 emissions using a synthesis of spatially representative forest inventory and supplemental plot data and life cycle assessment (LCA) of the wood removals. The specific objectives of this study were to:

1.

Determine the current net ecosystem production (NEP) and net biome production (NBP) of US West Coast Forests using observational datasets.

2.

Determine the net carbon emissions of forests for current (BAU) and for potential bioenergy management, accounting for in-situ and offsite carbon sources and sinks.

Chapter 3 evaluates the use of CLM4 in forests of the Pacific Northwest utilizing independent datasets with different temporal resolutions. The spatially explicit maps of

5 annual carbon stocks and net primary production (NPP) from the dataset compiled in

Chapter 1 were used to evaluate the landscape patterns of carbon stocks and fluxes. For seasonal evaluation, observed monthly gross photosynthesis (GPP) from AmeriFlux sites in the region (Krishnan et al. 2009, Thomas et al. 2009) were downloaded from the

FLUXNET website ( http://fluxnet.ornl.gov/ ) and compared to modeled monthly GPP.

The specific objective of this study was to evaluate the ability of CLM4 model to simulate historical and current carbon and nitrogen dynamics in both standard and modified format.

Chapter 4 uses the fully calibrated CLM4 model to examine the long term effects of varying bioenergy management strategies bioenergy on carbon and nitrogen dynamics under changing environmental conditions. We examined the following research questions:

1.

What are the interactive effects of changing climate, increasing atmospheric

CO

2

, N deposition and land use change on net ecosystem production (NEP)?

2.

How do the varying management scenarios affect carbon stocks and fluxes at different spatial scales?

3.

Does bioenergy management increase or decrease net CO

2

emissions to the atmosphere compared to BAU by the end of the 21 st

century?

4.

For which ecoregions, does net CO

2

emissions decrease compared to BAU for bioenergy management?

References

Birdsey, R. A. 2007. North American forests in The First State of the Carbon Cycle

Report (SOCCR): The North American Carbon Budget and Implications for the

Global Carbon Cycle. U.S. Climate Change Science Program and the

Subcommittee on Global Change Research, Washington, DC.

Breshears, D. D., O. B. Myers, C. W. Meyer, F. J. Barnes, C. B. Zou, C. D. Allen, N. G.

McDowell, and W. T. Pockman. 2008. Tree die-off in response to global changetype drought: mortality insights from a decade of plant water potential measurements. Frontiers in Ecology and the Environment 7:185-189.

Gurney, K. R., D. L. Mendoza, Y. Zhou, M. L. Fischer, C. C. Miller, S. Geethakumar,

6 and S. de la Rue du Can. 2009. High Resolution Fossil Fuel Combustion CO2

Emission Fluxes for the United States. Environmental Science & Technology

43:5535-5541.

Hayes, D. J., D. P. Turner, G. Stinson, A. D. McGuire, Y. Wei, T. O. West, L. S. Heath,

B. de Jong, B. G. McConkey, R. A. Birdsey, W. A. Kurz, A. R. Jacobson, D. N.

Huntzinger, Y. Pan, W. M. Post, and R. B. Cook. 2012. Reconciling estimates of the contemporary North American carbon balance among terrestrial biosphere models, atmospheric inversions, and a new approach for estimating net ecosystem exchange from inventory-based data. Global Change Biology 18:1282-1299.

Krishnan, P., T. A. Black, R. S. Jassal, B. Chen, and Z. Nesic. 2009. Interannual variability of the carbon balance of three different-aged Douglas-fir stands in the

Pacific Northwest. J. Geophys. Res. 114:G04011.

Lawrence, D., K. W. Oleson, M. G. Flanner, P. E. Thorton, S. C. Swenson, P. J.

Lawrence, X. Zeng, Z.-L. Yang, S. Levis, K. Skaguchi, G. B. Bonan, and A. G.

Slater. 2011. Parameterization Improvements and Functional and Structural

Advances in Version 4 of the Community Land Model. J. Adv. Model. Earth Syst.

3:27.

Lippke, B., E. Oneil, R. Harrison, K. Skog, L. Gustavsson, and R. Sathre. 2011. Life cycle impacts of forest management and wood utilization on carbon mitigation: knowns and unknowns. Carbon Management 2:303-333.

Lippke, B., J. Wilson, J. Meil, and A. Taylor. 2010. Characterizing the importance of carbon stored in wood products. Wood and Fiber Science 42:5-14.

Malmsheimer, R. W., J. L. Bowyer, J. S. Fried, E. Gee, R. L. Izlar, R. A. Miner, I. A.

Munn, E. Oneil, and W. C. Stewart. 2011. Managing Forests because Carbon

Matters: Integrating Energy, Products, and Land Management Policy. Journal of

Forestry 109:S7-S51.

McDowell, N. G., C. D. Allen, and L. Marshall. 2010. Growth, carbon-isotope discrimination, and drought-associated mortality across a Pinus ponderosa elevational transect. Global Change Biology 16:399-415.

Mitchell, S. R., M. E. Harmon, and K. E. B. O'Connell. 2012. Carbon debt and carbon sequestration parity in forest bioenergy production. GCB Bioenergy:n/a-n/a.

Raffa, K. F., B. H. Aukema, B. J. Bentz, A. L. Carroll, J. A. Hicke, M. G. Turner, and W.

H. Romme. 2008. Cross-scale Drivers of Natural Disturbances Prone to

Anthropogenic Amplification: The Dynamics of Bark Beetle Eruptions.

7

Bioscience 58:501-517.

Raymond, C. L., and D. L. Peterson. 2005. Fuel treatments alter the effects of wildfire in a mixed-evergreen forest, Oregon, USA. Canadian Journal of Forest Research

35:2981-2995.

Routa, J., S. KellomÄKi, A. KilpelÄInen, H. Peltola, and H. Strandman. 2011. Effects of forest management on the carbon dioxide emissions of wood energy in integrated production of timber and energy biomass. GCB Bioenergy 3:483-497.

Siemers, E. 2011. Northwest Energy plans $300M Oregon biomass investment. Pages http://www.sustainablebusinessoregon.com/articles/2011/2003/northwest-energyplans-2300m-oregon.html in Portland Business Journal, Portland.

Solomon, S., G.-K. Plattner, R. Knutti, and P. Friedlingstein. 2009. Irreversible climate change due to carbon dioxide emissions. Proceedings of the National Academy of

Sciences.

Stephens, S. L., J. J. Moghaddas, C. Edminster, C. E. Fiedler, S. Haase, M. Harrington, J.

E. Keeley, E. E. Knapp, J. D. McIver, K. Metlen, C. N. Skinner, and A.

Youngblood. 2009. Fire treatment effects on vegetation structure, fuels, and potential fire severity in western US forests. Ecological Applications 19:305-320.

Thomas, C. K., B. E. Law, J. Irvine, J. G. Martin, J. C. Pettijohn, and K. J. Davis. 2009.

Seasonal hydrology explains interannual and seasonal variation in carbon and water exchange in a semiarid mature ponderosa pine forest in central Oregon. J.

Geophys. Res. 114:G04006.

Turner, D. P., D. Ritts, B. E. Law, W. Cohen, Z. Yang, T. Hudiburg, J. L. Campbell, and

M. Duane. 2007. Scaling net ecosystem production and net biome production over a heterogeneous region in the western United States. Biogeosciences 4:597-612.

van Mantgem, P. J., N. L. Stephenson, J. C. Byrne, L. D. Daniels, J. F. Franklin, P. Z.

Fule, M. E. Harmon, A. J. Larson, J. M. Smith, A. H. Taylor, and T. T. Veblen.

2009. Widespread Increase of Tree Mortality Rates in the Western United States.

Science 323:521-524.

8

Waring, R. H., and J. F. Franklin. 1979. Evergreen Coniferous Forests of the Pacific

Northwest. Science 204:1380-1386.

Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. Warming and

Earlier Spring Increase Western U.S. Forest Wildfire Activity. Science 313:940-

943.

Williams, C. A., G. J. Collatz, J. Masek, and S. N. Goward. 2012. Carbon consequences of forest disturbance and recovery across the conterminous United States. Global

Biogeochem. Cycles 26:GB1005.

Winford, E., and J. Gaither Jr. In Press. Carbon outcomes from fuels treatment and bioenergy production in a Sierra Nevada forest Forest Ecology and Management.

REGIONAL CO

2

IMPLICATIONS OF FOREST BIOENERGY PRODUCTION

9

Authors: Tara W. Hudiburg, Beverly E. Law, Christian Wirth and Sebastiaan Luyssaert

Nature Climate Change

Nature Publishing Group, 4 Crinan Street, London N1 9XW UK

Volume 1: Issue 8

Chapter 2: Regional CO

2

implications of forest bioenergy production

Summary

Mitigation strategies for reducing CO

2

emissions include substitution of fossil fuel with bioenergy from forests (Buyx and Tait 2011), where carbon emitted is expected to be recaptured in the growth of new biomass to achieve zero net emissions (Gustavsson et al.

1995), and forest thinning to reduce wildfire emissions (Hurteau and North 2010). Here

10 we use forest inventory data to show that fire prevention measures and large-scale bioenergy harvest in US West Coast forests lead to 2-14% (46-405 Tg C) higher emissions compared to current management practices over the next 20 years. We studied

80 forest types in 19 ecoregions, and found that the current carbon sink in 16 of these ecoregions is sufficiently strong that it cannot be matched or exceeded through substitution of fossil fuels by forest bioenergy. If the sink in these ecoregions weakens below its current level by 30-60 g C m

-2

yr

-1

due to insect infestations, increased fire emissions, or reduced primary production, management schemes including bioenergy production may succeed in jointly reducing fire risk and carbon emissions. In the remaining three ecoregions, immediate implementation of fire prevention and biofuel policies may yield net emission savings. Hence, forest policy should consider current forest carbon balance, local forest conditions and ecosystem sustainability in establishing how to decrease emissions.

Main text

Policies are being developed worldwide to increase bioenergy production as a substitution for fossil fuel to mitigate fossil fuel-derived carbon dioxide emissions, the main cause of anthropogenic global climate change (Fargione et al. 2008, Richter Jr et al.

2009). However, the capacity for forest sector bioenergy production to offset carbon dioxide emissions is limited by fossil fuel emissions from this activity (harvest, transport, and manufacturing of wood products), and the lower energy output per unit carbon emitted compared with fossil fuels (Law and Harmon 2011). In addition, forest carbon sequestration can take decades to centuries to return to pre-harvest levels, depending on

11 initial conditions and amount of wood removed (Harmon and Marks 2002). The effects of changes in management on CO

2

emissions need to be evaluated against this baseline.

Consequently, energy policy implemented without full carbon accounting and understanding of the underlying processes risks increasing rather than decreasing emissions (Fargione et al. 2008, Searchinger et al. 2009).

In North America, there is increasing interest in partially meeting energy demands through large-scale forest thinning (Richter Jr et al. 2009) with the added benefit of preventing catastrophic wildfire and concurrent carbon loss (Hurteau and North 2010).

Although forest thinning can be economically feasible, sustainable, and an effective strategy for preventing wildfire where risk is high (Huggett Jr et al. 2008, Evans and

Finkral 2009), it remains unresolved whether this type of forest treatment can satisfy both the aims of preventing wildfire and reducing regional greenhouse gas emissions.

For both aims to be satisfied, it needs to be shown that: (1) reduction in carbon stocks due to thinning and the associated emissions are offset by avoiding fire emissions and substituting fossil fuel emissions with forest bioenergy, (2) the change in management results in less CO

2

emissions than the current or ‘baseline’ emissions, and

(3) short-term emission changes are sustained in the long-term. Determination of baseline forest sector carbon emissions can be accomplished by combining forest inventory data and life cycle assessment (LCA) that includes full carbon accounting of net biome production (NBP) on the land in addition to carbon emissions from bioenergy production and storage in wood products (LCA; (Law and Harmon 2011)). NBP is the annual net change of land-based forest carbon (NEP; photosynthesis minus respiration) after accounting for harvest removals and fire emissions.

Our study focused on the US West Coast (Washington, Oregon, and California), a diverse region due to the strong climatic gradient from the coast inland (300 – 2500 mm precipitation per year) and a total of 80 associated forest types ranging from temperate rainforests to semi-arid woodlands (Appendix A Table A1). The region is divided into 19 distinct ecoregions (Omernik 1987) based on climate, soil, and species characteristics, and includes a broad range of productivity, age structures, fire regimes and topography.

Mean net primary production (NPP) of the forest types range from 100 to 900 g C m

-2

yr

-1

(this study), falling within the global range of 100 to 1600 g C m

-2

yr

-1

reported for

12 temperate and boreal forests (Luyssaert et al. 2008). Forest land ownership is divided fairly evenly between public and private sectors having different management histories and objectives that affect forest carbon dynamics (Hudiburg et al. 2009).

Carbon sequestration rates vary greatly across the region, with mean NEP ranging from -85 g C m

-2

yr

-1

in the dry Northern Basin to over 400 g C m

-2

yr

-1

in the mesic

Coast Range. After accounting for fire emissions and substantial harvest removals, regional NBP remains a significant sink of 26±3 Tg C yr

-1

or 76±9 g C m

2

yr

-1

, similar to the US average (Birdsey 2007) and estimates for the member states of the European

Union (Luyssaert et al. 2009). Sixteen of the 19 ecoregions representing 98% of the forest area in the region are estimated to be carbon sinks (Fig 2.1a; exceptions are drier ecoregions where annual productivity is low and fire emissions are relatively high). Thus, the observed regional sink is not solely due to the region's highly productive rainforests that occupy 15% of the area. Within the region, California’s NBP is higher than that of

Oregon and Washington (107 versus 53-61 g C m

-2

yr

-1

), primarily due to differences in

NEP (Appendix A Table A2) and harvest between similar forest types within the same ecoregions that cross state boundaries (Appendix A Discussion and Table A3).

In addition to current management or Business-As-Usual (BAU, characterized by current preventive thinning and harvest levels), we designed three treatments (Appendix

A Fig A1a) to reflect the varying objectives of potential forest management systems: forest fire prevention by emphasizing removal of fuel ladders ('Fire Prevention') in fireprone areas, making fuel ladder removal economically feasible by emphasizing removal of additional marketable wood in fire–prone areas ('Economically Feasible'), or thinning all forestland regardless of fire risk to support energy production while contributing to fire prevention ('Bioenergy Production'). Removals are in addition to current harvest levels and are performed over a twenty year period such that 5% of the landscape is treated each year. Our reliance on a data-driven approach versus model simulations strengthens our analysis in the short-term, but limits our ability to make longer term predictions. Extending our study beyond a 20-year timeframe would over-stretch data use

because current forest growth is unlikely to represent future growth due to changes in climate, climate-related disturbance, and land use (Battles et al. 2008, Ryan 2010).

In our study region, we found that thinning reduced NBP under all three treatment scenarios for 13 of the 19 ecoregions representing 90% of the region's forest area. The

13 exceptions where NBP was not reduced were primarily due to high initial fire emissions compared to NEP (i.e. Northern Basin and North Cascades; Appendix A Fig A2). The dominant trend at the ecoregion level was mirrored at the regional level, with the

Bioenergy Production scenario (highest thinning level) resulting in the region becoming a net carbon source (Appendix A Table A2 and discussion of state-level estimates).

Regionally, forest biomass removals exceeded the potential losses from forest fires, reducing the in-situ forest carbon sink even after accounting for regrowth, as found in previous studies with different approaches or areas of inference (Mitchell et al. 2009,

Searchinger et al. 2009). Because we have assumed high reductions in fire emissions for the areas treated in each scenario, it is unlikely we are underestimating the benefit of preventive thinning on NBP.

It is important to recognize that even if the land-based flux is positive (a source) or zero (carbon neutral), decreases in NBP from BAU can increase CO

2

emissions to the atmosphere. LCA was used to estimate the net emissions of carbon to the atmosphere in each treatment scenario (Appendix A Fig A1b and Appendix A Tables A4 and A5). LCA at the ecoregion level revealed that emissions are increased for 10 out of 19 of the ecoregions (Fig 2.2), representing 80% of the forest area in the region. The combination of in-situ and wood-use carbon sinks and sources emit an additional 46, 181 and 405 Tg

C to the atmosphere over a 20 year period (2-14% increase), above that of the BAU forest management scenario for the Fire Prevention, Economically Feasible, and Bioenergy

Production treatments, respectively (Fig 2.3).

Sensitivity analysis of our results to a range of fire emission reductions, energy conversion efficiencies, wood product decomposition rates, and inclusion of wood substitution showed that carbon emissions varied -10 to 28% of the optimum values across the scenarios, depending on the combination of assumptions (Appendix A

Discussion and Table A6). The analysis revealed that an increase in estimated current fire

14 emissions (which effectively reduces the baseline sink) may decrease total atmospheric C emissions in the Fire Prevention scenario, but only given optimum conditions for all of the other parameters (e.g. 100% energy efficiency). Nevertheless, if fire frequency and intensity increase in the future (Rogers et al. 2011), emissions savings via forest bioenergy production may become possible, especially in ecoregions where the sink is already weak.

Previous case studies showed that harvesting an old-growth forest in the Pacific

Northwest (Harmon et al. 1990) or increasing the thinning removals of temperate forests is likely to deteriorate the forest and wood product carbon stock (Nunery and Keeton

2010). However, these studies were limited to a handful of sites, relied primarily on modeled results (Mitchell et al. 2009, Hurteau and North 2010) and did not account for the energy requirements of forest management and wood processing nor for the potential to substitute fossil fuels with bioenergy. We build on these results by including all ecoregions, all age classes (not just old-growth), three treatments including bioenergy production, and sector-based LCA. We found that even though forest sector emissions are compensated for by emission savings from bioenergy use, fewer forest fires, and wood product substitution, the end result is an increase in regional CO

2

emissions compared to

BAU as long as the regional sink persists.

To determine a threshold NBP for which bioenergy management reduces atmospheric CO

2

emissions compared to BAU, we applied the same assumptions used in the LCA. We found that if the NBP drops by 50-60 g C m

-2

yr

-1

in currently productive ecoregions or 15-30 g C m

-2

yr

-1

in currently less productive ecoregions, bioenergy management would come with CO

2

emissions savings compared to BAU (Fig 2.1c).

Aggregating the ecoregion thresholds translate into a regional mean NBP of 45 g C m

-2 yr

-1 or a 41% reduction on average. Reductions in NBP may occur due to increased mortality and/or decreased growth due to climate, fire, or insect outbreaks. However reductions in NBP from increased harvest does not qualify because harvest increases emissions; wood carbon enters the products/bioenergy chain where subsequent losses occur. We cannot predict from the data when the threshold NBP would occur because a high resolution process-based model with the ability to incorporate future climate,

nitrogen deposition, age dynamics, disturbance, and management would need to be utilized, which is beyond the scope of this study.

Ecoregion threshold NBP is dependent on the scenario treatment removals and area because the Fire Prevention treatment targets only those areas most likely to burn.

For example, to reduce emissions in the Sierra Nevada, baseline NBP would have to decrease by as much as 84 for the Bioenergy Production scenario versus only 13 g C m

-2 yr

-1

for the Fire Prevention scenario. In ecoregions where current sinks are marginal or

15 weakened by climate, fire, or insect outbreaks there may be a combination of harvest intensity and bioenergy production that reduces forest sector emissions. In 9 of the ecoregions where forests are carbon neutral or a source of CO

2 to the atmosphere and/or fire emissions are high for BAU, total CO

2

emissions under the Fire Prevention scenario could be reduced compared to BAU. They provide examples where management strategies for carbon emission reduction or sequestration should differ from the majority of the region; a one-size-fits-all approach will not work (Marland and Schlamadinger

1997). Finally, large areas in the Northern Rockies (i.e. Colorado and Wyoming) are currently experiencing increases in forest mortality due to beetle-kill, a trend which could continue in a warmer climate (Evangelista et al. 2011). These areas may already be at or below the threshold NBP; if so, they could benefit from targeted bioenergy implementation. However, simply lowering current regional harvest intensities in areas where NBP is not weakened also reduces emissions (Appendix A Discussion and Fig

A3). Also, as we have assumed large-scale implementation of these strategies in addition to BAU harvest, we may be overestimating future harvest even though harvest has declined significantly since 1990 due to restrictions placed on harvest on federal lands as part of the Northwest Forest Plan. If the strategies were used to substitute for BAU harvest, the outcome on NBP would be much different (i.e. increased for the Fire

Prevention scenario).

Our study is one of the first to provide a full carbon accounting, including all of the sinks and sources of carbon emissions from the forestry sector and the current in-situ sink for such a large area. Given the diversity of woody ecosystems in the study region ranging from highly productive temperate rainforests to less productive semi-arid

woodlands, the trends in response likely apply to other temperate regions globally

(Appendix A Table A1) where forests are currently a strong net carbon-sink (i.e. Eastern

16

US, China and Europe), although the extent of the effect remains to be established.

Greenhouse gas reduction plans call for up to 10% reductions in emissions by

2020 and forest-derived fuels are being proposed as a carbon-neutral solution to reducing energy emissions. In all of our proposed scenarios, increases in harvest volume on the

US West Coast will on average result in regional emission increases above current levels, although there are a few ecoregions where the tested scenarios could result in emission savings. As long as the current in-situ NBP persists, increasing harvest volumes in support of bioenergy production is counterproductive for reducing CO

2 emissions. In this study region, the current in-situ NBP in tree biomass, woody detritus, and soil carbon is more beneficial in contributing to reduction of anthropogenic carbon dioxide emissions than increasing harvest to substitute fossil fuels with bioenergy from forests.

Although large uncertainty remains for regional forecasts to year 2050 or 2100, it is expected that forest carbon sinks will diminish over time because of aging of the forests, saturation of the CO

2

-fertilization and N-deposition effects, and increased mortality due to climate or insects (van Mantgem et al. 2009, Stinson et al. 2011). This would require new assessments to identify management options appropriate for each situation. Carbon-management is not the sole criteria that should be considered when planning forest management. Our findings should thus also be evaluated against other ecosystem services such as habitat, genetic and species diversity, watershed protection, and natural adaptation to climate change.

Methods Summary

We quantified forest sequestration rates and test forest thinning scenarios across the region using a data-intensive approach which, for the first time, takes into account the diversity of forest characteristics and management. We combined Landsat remote sensing data with inventories and ancillary data to map current forest NEP, NBP, and changes in

NBP with three thinning scenarios. The approach can be applied at multiple scales of analysis in other regions.

17

We combined spatially representative observational data from over 6000 FIA plots (see Appendix A Methods; Appendix A Table A7 ) with remote sensing products on forest type, age, and fire risk (USGS 2009), a global data compilation of wood decomposition data and 200 supplemental plots (Hudiburg et al. 2009) to provide new estimates of US West Coast (~34 million hectares) forest biomass carbon stocks

(Appendix A Table A8), net ecosystem production (NEP, the balance of photosynthesis and respiration), and net biome production (NBP, the in-situ net forest carbon-sink accounting for removals). We included all forestland in our analysis across all age classes

(20-800 years old) and management regimes. Plot values were aggregated by climatic region (ecoregion), age class, and forest type and this look-up table was used to assign a value to each associated 30 meter pixel.

We use regional combustion coefficients to determine fire emissions. Only 3-8% of live tree biomass is actually combusted and emitted in high severity fire in the Pacific

Northwest (Campbell et al., 2007), contrary to other studies that report much higher emissions because they assume 30% of all aboveground woody biomass is consumed

(Wiedinmyer and Hurteau 2010). Although the latter contradicts extensive field observations (Campbell et al. 2007, Meigs et al. 2009) and modeling studies (Ottmar et al. 2006) in the region, we included 30% as the upper-end combustion factor in our sensitivity analysis (Appendix Table A9).

In addition to the spatially explicit estimates of stocks and fluxes under current management or Business-As-Usual (BAU, current forest harvest), three treatments were designed (Fire Prevention, Economically Feasible, and Bioenergy Production; Appendix

A Fig A1a) to reflect the varying objectives of potential future forest management over the next 20 years, within the proposed time period for CO

2

reductions in the U.S. Areas were prioritized for treatment by fire risk and frequency. The proposed treatments result in additional harvest removals because we assume the current harvest rate for wood products will continue in the future. We limit our specific analysis to the short term because this is the timeframe suitable for policymakers, effectiveness of fire protection treatments, and an appropriate use of the data-driven approach. However, to investigate conditions (i.e. sink saturation) that could invalidate our short term results in the long-

term, we also calculated the in situ NBP at which the atmosphere may benefit from bioenergy removals.

Lastly, we studied the net effects of the thinning treatments on atmospheric CO

2 by a life cycle assessment (LCA) of carbon sources and sinks that includes the post-

18 thinning NBP, and wood use (harvest, transport, manufacturing, decomposition, wood product substitution, conversion and use of bioenergy, and displacement of fossil fuel extraction emissions; Appendix A Fig A1b and Appendix A Table A4 and A5).

References

Battles, J., T. Robards, A. Das, K. Waring, J. Gilless, G. Biging, and F. Schurr. 2008.

Climate change impacts on forest growth and tree mortality: a data-driven modeling study in the mixed-conifer forest of the Sierra Nevada, California.

Climatic Change 87:193-213.

Birdsey, R. A. 2007. North American forests in The First State of the Carbon Cycle

Report (SOCCR): The North American Carbon Budget and Implications for the

Global Carbon Cycle. U.S. Climate Change Science Program and the

Subcommittee on Global Change Research, Washington, DC.

Bonan, G. B. 2008. Forests and Climate Change: Forcings, Feedbacks, and the Climate

Benefits of Forests. Science 320:1444-1449.

Bonan, G. B., P. J. Lawrence, K. W. Oleson, S. Levis, M. Jung, M. Reichstein, D. M.

Lawrence, and S. C. Swenson. 2011. Improving canopy processes in the

Community Land Model version 4 (CLM4) using global flux fields empirically inferred from FLUXNET data. J. Geophys. Res. 116:G02014.

Buyx, A., and J. Tait. 2011. Ethical Framework for Biofuels. Science.

Campbell, J., D. Donato, D. Azuma, and B. Law. 2007. Pyrogenic carbon emission from a large wildfire in Oregon, United States. J. Geophys. Res. 112.

Evangelista, P. H., S. Kumar, T. J. Stohlgren, and N. E. Young. 2011. Assessing forest vulnerability and the potential distribution of pine beetles under current and future climate scenarios in the Interior West of the US. Forest Ecology and Management

262:307-316.

Evans, A., and A. Finkral. 2009. From renewable energy to fire risk reduction: a synthesis of biomass harvesting and utilization case studies in US forests. GCB

Bioenergy 1:211-219.

Fargione, J., J. Hill, D. Tilman, S. Polasky, and P. Hawthorne. 2008. Land clearing and

19 the biofuel carbon debt. Science 319:1235-1238.

Gustavsson, L., P. l. Börjesson, B. Johansson, and P. Svenningsson. 1995. Reducing

CO2 emissions by substituting biomass for fossil fuels. Energy 20:1097-1113.

Harmon, M. E., W. K. Ferrell, and J. F. Franklin. 1990. Effects on carbon storage of conversion of old-growth forests to young forests. Science 247:699-702.

Harmon, M. E., and B. Marks. 2002. Effects of silvicultural practices on carbon stores in

Douglas-fir – western hemlock forests in the Pacific Northwest, U.S.A.: results from a simulation model. Canadian Journal of Forest Research/Revue Canadienne de Recherche Forestiere 32:863-877.

Hudiburg, T., B. E. Law, D. P. Turner, J. Campbell, D. Donato, and M. Duane. 2009.

Carbon dynamics of Oregon and Northern California forests and potential landbased carbon storage. Ecological Applications 19:163-180.

Hudiburg, T. W., B. E. Law, C. Wirth, and S. Luyssaert. 2011. Regional carbon dioxide implications of forest bioenergy production. Nature Clim. Change 1:419-423.

Huggett Jr, R. J., K. L. Abt, and W. Shepperd. 2008. Efficacy of mechanical fuel treatments for reducing wildfire hazard. Forest Policy and Economics 10:408-414.

Hurteau, M. D., and M. North. 2010. Carbon recovery rates following different wildfire risk mitigation treatments. Forest Ecology and Management 260:930-937.

Law, B. E., and M. E. Harmon. 2011. Forest sector carbon management, measurement and verification, and discussion of policy related to climate change. Carbon

Management 2:73-84.

Luyssaert, S., P. Ciais, S. L. Piao, E.-D. Schulze, M. Jung, S. Zaehle, M. J. Schelhaas, M.

Reichstein, G. Churkina, D. Papale, G. Abril, C. Beer, J. Grace, D. Loustau, G.

Matteucci, F. Magnani, G. J. Nabuurs, H. Verbeeck, M. Sulkava, G. R. v. d.

WERF, and I. A. Janssens. 2009. The European carbon balance: part 3: Forests.

Global Change Biology 16:1429-1450.

Luyssaert, S., E. D. Schulze, A. Borner, A. Knohl, D. Hessenmoller, B. E. Law, P. Ciais, and J. Grace. 2008. Old-growth forests as global carbon sinks. Nature 455:213-

215.

Marland, G., and B. Schlamadinger. 1997. Forests for carbon sequestration or fossil fuel substitution? A sensitivity analysis. Biomass and Bioenergy 13:389-397.

Meigs, G., D. Donato, J. Campbell, J. Martin, and B. Law. 2009. Forest fire impacts on

20 carbon uptake, storage, and emission: the role of burn severity in the Eastern

Cascades, Oregon. Ecosystems 12:1246-1267.

Mitchell, S. R., M. E. Harmon, and K. E. B. O'Connel. 2009. Forest fuel reduction alters fire severity and long-term carbon storage in three Pacific Northwest ecosystems.

Ecological Applications 19.

Mitchell, S. R., M. E. Harmon, and K. E. B. O'Connell. 2012. Carbon debt and carbon sequestration parity in forest bioenergy production. GCB Bioenergy:n/a-n/a.

Nunery, J. S., and W. S. Keeton. 2010. Forest carbon storage in the northeastern United

States: Net effects of harvesting frequency, post-harvest retention, and wood products. Forest Ecology and Management 259:1363-1375.

Omernik, J. M. 1987. Ecoregions of the conterminous United States. Map (scale

1:7,500,000). Annals of the Association of American Geographers 77:118-125.

Ottmar, R. D., S. J. Pritchard, R. E. Vihnanek, and D. V. Sandberg. 2006. Modification and Validation of Fuel Consumption Models for Shrub and Forested lands in the

Southwest, Pacific Northwest, Rockies,Midwest, Southeast and Alaska, Final

Report JFSP Project 98-1-9-06.

Richter Jr, D., D. H. Jenkins, J. T. Karakash, J. Knight, L. R. McCreery, and K. P.

Nemestothy. 2009. Resource policy: wood energy in America. Science 323:1432-

1433.

Rogers, B., R. Neilson, R. Drapek, J. Lenihan, J. Wells, D. Bachelet, and B. Law. 2011.

Impacts of climate change on fire regimes and carbon stocks of the U.S. Pacific

Northwest. J. Geophys. Res. 116 G03037.

Ryan, M. G. 2010. Temperature and tree growth. Tree Physiology 30:667-668.

Searchinger, T. D., S. P. Hamburg, J. Melillo, W. Chameides, P. Havlik, D. M. Kammen,

G. E. Likens, R. N. Lubowski, M. Obersteiner, M. Oppenheimer, G. Philip

Robertson, W. H. Schlesinger, and G. David Tilman. 2009. Fixing a critical climate accounting error. Science 326:527-528.

Stinson, G., W. A. Kurz, C. E. Smyth, E. T. Neilson, C. C. Dymond, J. M. Metsaranta, C.

Boisvenue, G. J. Rampley, Q. Li, T. M. White, and D. Blain. 2011. An inventorybased analysis of Canada's managed forest carbon dynamics, 1990 to 2008.

Global Change Biology 17:2227-2244.

21

Thornton, P. E., and N. E. Zimmermann. 2007. An improved canopy integration scheme for a land surface model with prognostic canopy structure. Journal of Cllimate

20:3902-3923.

USGS. 2009. The National Map LANDFIRE: LANDFIRE National Existing Vegetation

Type layer. Pages http://gisdata.usgs.net/website/landfire/ in U.S. Department of

Interior, Geological Survey. U.S. Department of Interior, Geological Survey. van Mantgem, P. J., N. L. Stephenson, J. C. Byrne, L. D. Daniels, J. F. Franklin, P. Z.

Fule, M. E. Harmon, A. J. Larson, J. M. Smith, A. H. Taylor, and T. T. Veblen.

2009. Widespread Increase of Tree Mortality Rates in the Western United States.

Science 323:521-524.

Wiedinmyer, C., and M. D. Hurteau. 2010. Prescribed fire as a means of reducing forest carbon emissions in the western United States. Environmental Science &

Technology 44:1926-1932.

22

Figure 2.1. Current net biome productivity (NBP), NBP uncertainty, and change in NBP required for zero net emissions compared to business-as-usual (BAU). Positive values indicate forest sinks while negative values are carbon sources to the atmosphere.

Uncertainty estimates (b) were calculated using Monte Carlo simulations of mean forest type values for the components of NBP (net ecosystem productivity, fire, and harvest) combined with the uncertainty associated with remote sensing land cover estimates.

Change in NBP (c) represents the amount NBP would need to decrease to reach a threshold NPB where bioenergy management may result in emission decreases to the atmosphere.

23

Figure 2.2. Life Cycle Assessment carbon emission trends by ecoregion for each scenario. The x-axis is the total harvest (BAU + treatment) and the y-axis is the total CO

2 flux in Tg C yr

-1

for each ecoregion. Colored circles represent each scenario (Green =

BAU, Yellow = Fire Prevention, Orange = Economically Feasible, and Red = Bioenergy

Production). Grey circles are the values for each sensitivity analysis set of parameters and the error bars represent the estimate uncertainty. For most ecoregions, the treatments increase emissions to the atmosphere.

Figure 2.3. Total regional carbon sinks, sources, and added emissions for each scenario.

Units are in Tg C yr

-1

. Life cycle assessment estimates account for changes in carbon on land in addition to emissions associated with production, transport and usage of wood and substitution and displacement of fossil fuel emissions associated with use and extraction.

BAU results in the lowest anthropogenic emissions from the forest sector.

24

Chapter 3: Evaluation and improvement of the Community Land Model (CLM 4.0) in Pacific Northwest Forests

Abstract

25

Ecosystem process models are important tools for determining the interactive effects of global change and disturbance on forest carbon dynamics. Here we evaluated and improved terrestrial carbon and water cycling simulated by the Community Land Model

(CLM4), the land model portion of the Community Earth System Model (CESM1.0.4).

Our analysis was conducted in Pacific Northwest forests using AmeriFlux and inventory data for the period 2001-2006. We compared simulated variables with a spatially representative dataset of forest carbon stocks, and net primary production (NPP; federal forest inventory data; FIA), and compared simulated gross primary production (GPP) with AmeriFlux eddy-covariance tower data at wet and dry sites in the region. In addition to evaluation of model uncertainty, we tested the use of a downscaled regional forcing dataset at 1/8 th

degree (15 km) resolution versus the standard 1-2 degree (100-200km) forcing dataset supplied for the purpose of global model simulations. We improved model estimates by making modifications to CLM4 to allow physiological parameters

(e.g. foliage carbon to nitrogen ratios and specific leaf area), mortality, and wood allocation to vary spatially within a plant functional type (PFT). Prior to modifications, default parameters resulted in underestimation of stem biomass in all forested ecoregions except the Blue Mountains and annual NPP was both over and underestimated. After modifications, model estimates of mean NPP fell within the observed range of uncertainty in all ecoregions (two-sided p value = 0.8) and the underestimation of stem biomass was reduced. At the tower sites, modeled annual GPP fell within the observed range of uncertainty (reduced chi-square statistic < 1) at both sites, however summer GPP was consistently underestimated and did not fall within the observed range of uncertainty.

Modeled annual NPP was underestimated by an average of 24% when compared to biometric estimates of NPP at the Metolius site while modeled GPP was underestimated by an average of 30% compared to the eddy-covariance data. However, the ratio of

NPP/GPP was nearly identical for both equaling 37% (observed) and 36% (modeled).

26

The low bias in summer GPP could be due to several possible reasons including variation in seasonal V cmax

response to temperature, precipitation, and nitrogen dynamics. We found improved parameterization of foliar nitrogen content and nitrogen availability increased monthly GPP in all months resulting in correct annual sums of GPP, but did not improve seasonal dynamics. This indicates that the model algorithm controlling V cmax response to a variety of other factors including temperature, daylength, and / or soil water content may need revising.

Introduction

Modeling and understanding the response of terrestrial ecosystems to changing environmental conditions and land use change are primary goals of climate mitigation policy (IPCC 2007, Moss et al. 2010, NRC 2010a, Pacala 2010). The Intergovernmental

Panel for Climate Change (IPCC) synthesizes estimates of future climate change impacts on terrestrial carbon cycling through the use of a specific set of global circulation models

(IPCC 2007). Among them is the Community Earth System Model (CESM) of the

National Center for Atmospheric Research (NCAR). The land model component

(Community Land Model; CLM) has been extensively used to evaluate and predict the net carbon uptake and loss from terrestrial biomes, particularly forests (Thornton and

Zimmermann 2007, Bonan et al. 2011).

Recent assessments of seasonal performance of multiple terrestrial biosphere models indicated better performance in forest ecosystems, particularly in evergreen forests, during the summer growing season than in other biomes and seasons (Schwalm et al. 2010, Schaefer et al. in press). The highest skill across biomes was in models that prescribed canopy phenology and did not use a daily time step (Schwalm et al. 2010). In an assessment with FLUXNET tower data, Schaefer et al. (in press) found that none of the models estimated gross photosynthesis (GPP) at all sites within observed uncertainty.

The most important factor influencing model performance was light-use efficiency computed from the slope of the GPP – light response curve. This is influenced by photosynthetic parameters, indicating better parameter values are needed for variables influencing LUE. The model evaluations also suggested more detailed assessments need

to be performed with individual models. The photosynthesis model of (Farquhar et al.

27

1980) is widely used in vegetation models, and an evaluation at more than 200

FLUXNET sites showed that simple PFT classification of photosynthesis parameters introduces uncertainty in photosynthesis and transpiration fluxes (Groenendijk et al.

2012b), and including V cmax

seasonality can improve predictions (Groenendijk et al.

2012a). The study suggested focusing on the effects of seasonal foliar N on Vcmax. In an assessment of CLM4 with FLUXNET data, Bonan et al. (2011) also found that the bias in annual GPP could be reduced by including improved estimates of photosynthetic parameters. Finally, CLM4 was also found to overestimate shade-leaf photosynthesis leading to overestimation of canopy GPP when the nitrogen limitation functionality was inactive (Bonan et al., 2012) suggesting a multi-layer canopy could improve initial GPP calculations before downregulation due to nitrogen limitation is imposed.

In this paper, we evaluate the Community Land Model (version 4.0) portion of the

Community Earth System Model ( CESM1.0.4

). CLM4 is the latest in a series of land models developed for the CESM and runs at a half-hourly timestep. CLM4 includes coupled carbon and nitrogen processes and examines the physical, chemical, and biological processes through which terrestrial ecosystems affect and are affected by climate across a variety of spatial and temporal scales. Recent model releases include improvements to hydrology and an integrated transient land cover and land use change dataset (Lawrence et al. 2011). The transient land cover dataset includes historical wood harvest which is known to have a large influence on Pacific Northwest forest carbon storage and uptake (Harmon et al. 1990, Law et al. 2004). When the carbon-nitrogen biogeochemistry is active, CLM4 uses prognostic canopy phenology to determine leaf and stem area index (LAI and SAI) and vegetation heights. Potential GPP is calculated from leaf photosynthetic rate without nitrogen constraint. Actual GPP is computed from nitrogen limitations to potential GPP.

Our objective was to evaluate carbon stocks and fluxes simulated by CLM4 in forests and woodlands of the Pacific Northwest using independent datasets with different temporal resolutions. Annual carbon stocks and net primary production (NPP) were computed from inventory data that had been scaled to the region with remote sensing data

to produce a spatial dataset (Hudiburg et al. 2011), and the seasonal data were gross photosynthesis (GPP) from two AmeriFlux sites in the region (Krishnan et al. 2009,

Thomas et al. 2009). Model evaluation with an observation-based spatially representative dataset such as federal forest inventory data (FIA) is critical in order to constrain model

28 estimates, facilitate model development and ultimately reduce model uncertainty.

Combined with seasonal data and analysis over a strong climatic gradient, diagnostics can point to areas for model improvements. The specific objectives of this study are to 1) evaluate regional model performance against spatially representative FIA estimates of stem wood biomass and annual NPP; 2) evaluate seasonal model performance against monthly tower GPP utilizing observed physiological parameters; and 3) examine the roles of modeled LUE, temperature, and nitrogen limitation in determining seasonal patterns of GPP. Temperature is a dominant control on the seasonal variation in GPP

(Schaefer et al. In Press)whereas nitrogen limitation is a major determinant of seasonal maximum GPP (Bonan et al. 2011).

The accuracy and uncertainty associated with regional estimates of carbon fluxes by CLM4 is unknown because model output is not usually validated against spatially representative observations. The federal forest inventory (FIA) collects data on an annual basis on all forestland regardless of ownership or location resulting in an intensively sampled landscape through which modeled estimates of forest carbon stocks and fluxes could be evaluated. FIA data provide measurements that can be used to calculate the carbon density of live and dead trees, woody detritus, and understory shrubs. Wood increment data are also provided, allowing estimation of 5-10 year average bole wood production, depending on the repeat measurement cycle. Combined with remote sensing land cover products, reliable maps of wood carbon stocks , net primary production (NPP)

, heterotrophic respiration (R h

), and net ecosystem production (NEP) can be produced

(Hudiburg et al. 2011, Law et al. 2012) to validate model output and identify model biases. The associated uncertainty in the FIA estimates can also be calculated providing a range of values or baseline conditions modeling activities should be constrained by before making predictions about future conditions.

Here, we evaluate CLM4 with FIA data maps of wood carbon stocks and NPP

29 averaged for the years 2001-2006 in Oregon forests. We improve on model estimates by calibrating with supplemental plot data collected at over 100 sites across the study area.

The region is characterized by tremendous climate variation and it has been recommended that a minimum of 15 km resolution be utilized in such heterogeneous terrain (Salathé et al. 2010). We utilize a new downscaled 200 year 1/8 th

degree (15 km) resolution dataset specifically designed for use in the Pacific Northwest (Salathe et al.

2007). Furthermore, we also test the model’s ability to determine the seasonal timing and magnitude of GPP by comparing modeled estimates with eddy-covariance data in the region, and evaluate tower annual NPP and GPP for consistency (Luyssaert et al. 2009).

Finally, we make modifications to model algorithms and plant functional type (PFT) physiological parameterization to improve model performance. By using existing datasets and uncertainty quantification, this research provides a solid scientific foundation for evaluating carbon dynamics under conditions of future climate change and land management practices at local and regional scales.

Methods: Model Description

The model used to simulate these processes is the Community Land Model

(CLM4) portion of the Community Earth System Model (CESM 1.0.4) of the National

Center for Atmospheric Research (Oleson et al. 2010). CLM4 uses hourly climate data, ecophysiological characteristics, site physical characteristics, and site history to estimate the daily fluxes of carbon, nitrogen, and water between the atmosphere, plant state variables, and litter and soil state variables. State variables are the live and dead carbon pools. CLM4 examines the physical, chemical, and biological processes through which terrestrial ecosystems affect and are affected by climate across a variety of spatial and temporal scales. The basic assumption is that terrestrial ecosystems, through their cycling of energy, water, chemical elements, and trace gases, are important determinants of climate. Model components consist of biogeophysics, hydrologic cycle, biogeochemistry and dynamic vegetation. The land surface is divided into five sub-grid land cover types

(glacier, lake, wetland, urban, vegetated) in each grid cell. The vegetated portion of a grid

30 cell is further divided into patches of plant functional types (PFTs), each with its own leaf and stem area index and canopy height. The multiple PFTs within the grid cell compete for water and nutrients on a single soil column.

Recent model improvements include updates to hydrology and an integrated transient land cover and land use change dataset that accounts for wood harvest. Updates to the hydrology include improved ground evaporation parameterization and snowpack heating and aging resulting in higher snow cover, cooler soil temperatures in organic-rich soils, greater river discharge, and lower albedos over forests and grasslands (Lawrence et al. 2011). The transient land use dataset has been formatted for use by CLM4 from a global historical transient land use and land cover change dataset (Hurtt et al. 2006) covering the period from 1850–2005. The dataset describes land cover and its change via four classes of vegetation (crop, pasture, primary vegetation, and secondary vegetation.)

The dataset also describes the annual fraction of land that is transformed from one PFT to another including the conversion of primary forest to secondary forest which is essential for studies attempting to track forest carbon storage and uptake over time. This dataset essentially provides a human-induced disturbance history of forests which can be used to test the model’s ability to simulate carbon dynamics following harvest.

Methods: Model calibration and forcing datasets

CLM4 has over 40 physiological parameters for each of the 17 plant functional types and a default constant mortality rate of 2% for all PFTs. The PFTs include 8 different forested and 3 different shrub land PFTs (Appendix B; Table B1 and B2). We use data collected on 100 supplemental plots located throughout the study region (Figure

3.1) to calibrate the physiology file and mortality rate. Default physiology values were used for parameters where data was unavailable and for parameters which are known to have little influence on biomass and NPP. Assessment and sensitivity analysis of CLM4 to the parameter values has been tested and described (White et al. 2000, Bonan et al.

2011, Lawrence et al. 2011) and we incorporate this knowledge to facilitate our regional calibration. The supplemental plot dataset includes measurements of specific leaf area

(SLA; projected), foliar carbon nitrogen ratios, litter carbon nitrogen ratios, and leaf

longevity for the major tree species across the study region. We use the PFT mean for

31 each of these parameters in the calibrated physiology file for default model configuration evaluation, heretofore indicated as ‘CLM4’ in figures and text. Prior modeling studies in this region indicate that a dynamic mortality rate that varies with age and/or disturbance type is necessary to predict the correct seasonal and annual carbon fluxes (Turner et al.

2007, Edburg et al. 2011), especially in the drier forest types where mortality decreases as a percentage of live biomass as stands age. Mortality rates in Oregon forests range from 0.5 to 2% (Turner et al. 2007, Hudiburg et al. 2009) in the absence of stand replacing disturbance. Since dynamic mortality algorithms have not been incorporated into CLM4, we chose to use 1.0% as a static mean value for the default configuration simulations. To improve on this, we parameterized CLM4_mod with PFT mortality rates within each ecoregion based on inventory data (Hudiburg et al. 2009). Ecoregions are areas within which ecosystems (and the type, quality, and quantity of environmental resources) are generally similar (Omernik 1987).

Offline simulations of CLM4 are typically forced with the NCEP reanalysis dataset (Qian et al. 2006) provided by NCAR. While this dataset includes the required climate and forcing variables at a sub-daily timestep, the resolution (~2 degree) is not adequate for regional simulations in Oregon. For this reason, we forced the model with a

1/8 th

degree regional dataset assembled by the Climate Impacts Group at the University of Washington ( http://www.cses.washington.edu/data/ipccar4/ ). Daily gridded historical

(1900-2000) and future (2000-2100) data are provided from three different models participating in the IPCC assessment reports. The regional downscaled dataset includes daily precipitation, minimum and maximum temperature, and wind speed. Because

CLM4 also requires shortwave radiation and relative humidity we calculated values incorporating algorithms from DAYMET (Thornton et al. 1997) and methods for subdaily calculations as described by (Appendix B;Göeckede et al. 2010).

Methods: Model simulations and evaluation

All model simulations are summarized in Table 3.1. Model spinups required 650 years to reach equilibrium conditions and were performed using regional physiology

parameters and the downscaled climate forcing dataset (PNW). The fire module was turned ‘off’ for spinup because in some ecoregions, biomass would ‘burn up’ before reaching equilibrium conditions. Based on previous experience with CLM4 in the region, fire combustion estimates were reduced by 50%. After spinup, control simulations were

32 run using the different model versions (see model development below). The control simulations were run from 1850-2006 using preindustrial CO

2

and N deposition levels and constant 1850 land cover. Transient simulations of each model version were run by changing CO

2

, nitrogen deposition, land use and land cover, and/or climate for the period from 1850-2006. Separate simulations allow for attributing the responses due to climate, land use change, and model versions. Transient CO

2

, nitrogen deposition (Ndep), and land cover files are annual files covering the period from 1850-2006. Ndep and land cover are spatially explicit datasets interpolated from a half-degree global dataset, while the CO

2

file is a single value for the entire region which changes annually.

Methods: Observations

We combined spatially representative observational data from over 3125 FIA plots measured during 2001 to 2006 with remote sensing products on forest cover, ecoregion, and age and a global data compilation of wood decomposition data (Wirth et al. 2010) to provide current maps of Oregon (~12 million hectares) forest biomass carbon stocks, heterotrophic respiration (R h

), net primary production (NPP), and net ecosystem production (NEP). We included all forestland in our analysis across all age classes (20-

800 years old) and management regimes. Plot values were aggregated by ecoregion, age class, and forest type and this look-up table was used to assign a value to each associated

30 meter pixel. Methods and uncertainty of the inventory biomass and carbon flux calculations are fully described in (Hudiburg et al. 2009, Hudiburg et al. 2011). Tree biomass is calculated using ecoregion- and species-specific allometric equations from diameter at breast height (DBH) and height. Woody detritus length and diameter are recorded along transects for each plot in the FIA dataset. Woody detritus biomass is then calculated from piece volume and decay class density. Wood NPP is calculated using increment data from wood cores collected on the FIA plots, foliage NPP is foliage

biomass divided by the average leaf retention time measured at the intensive plots, and

33 fine root NPP is calculated from fine root biomass multiplied by the average root turnover rates from intensive plot data (Appendix A, Table A8).

Monthly mean values of GPP from eddy-covariance tower data and the associated uncertainty for the period from 2001-2006 were downloaded from the AmeriFlux database for two tower sites in the region, one in a mesic forest of the Coast Range ecoregion (Campbell River, BC) and one in a dry forest of the East Cascades (Metolius,

OR). The provided uncertainty estimates for the flux tower observations were calculated for the model-data synthesis project by Schwalm et al. (2010). These sites represent mature secondary forest under private ownership (50- 80 years) with contrasting rates of biomass accumulation and NPP primarily due to climatic conditions (Law et al. 2003,

Law et al. 2004). Seasonal patterns of modeled GPP are compared with the flux tower observations.

Methods: Uncertainty analysis

We calculated the total uncertainty in the current FIA estimates using the propagation of error approach (Williams et al. 2012). The propagation of error approach is a method for combing error terms to calculate total uncertainty in an estimate. Monte

Carlo simulations were used to estimate the uncertainty due to variation in region- and/or species-specific allometry by using the mean and standard deviations for stem biomass and NPP calculated using three alternative sets of allometric equations. The full suite of species-specific equations that use tree diameter (DBH) and height (preferred) were compared to a DBH-only national set (Jenkins et al. 2003) and to a grouped forest type set. Total uncertainty in FIA estimates was calculated as the combined uncertainty of the allometric equations and land cover estimates (10%) using the propagation of error approach (NRC 2010b). Ecoregion level uncertainty estimates are represented in the figures by the grey error bars and in tables with ‘±’ symbols (NRC 2010a). For finer scale evaluation (i.e. grid cell size), the uncertainty in the observations was used to evaluate model performance using the reduced chi-square ( ) statistic (Schwalm et al. 2010) and

model bias defined as the mean of the model-data residuals for different groups of interest. Inventory plot data were grouped by ecoregion and then aggregated by precipitation regime, biome type, and age group within each ecoregion. Plots with annual precipitation greater than 800 mm were considered ‘Wet’ and less than 800 mm as ‘Dry’.

34

FIA forest types are either evergreen needleleaf (ENF), deciduous broadleaf (DBF), or mixed ENV/DBF (MXD). Complete analysis and statistics were performed for all biomes, however since 92% of the forested area in the study region is ENF, we include only the ENF biome in the figures and tables of this document. Age groups are defined as young (< 60 years) and mature (> 60 years).

The reduced is the squared sum of the residuals normalized by the observation uncertainty and divided by the total number of samples (i.e. the mean for an arbitrary group of interest):

Reduced )

2 where, is the uncertainty in the i th

observation, ‘2’ normalizes the uncertainty to correspond to a 95% confidence interval, and the summation is across the aggregated data groups within each ecoregion (Schwalm et al. 2010). Reduced values close to 1 indicate model-data unity or agreement. Model bias can be estimated as the mean of the residuals:

Bias = where i is the group of interest and positive values indicate average overestimation by the model and negative values indicate average underestimation.

Methods: Model development

The model sensitivity to specific parameter value ranges has been tested and described (White et al. 2000, Thornton et al. 2002).We aimed to reduce overall

uncertainly by calibrating with known regional mean values and evaluating with inventory data. This was in part due to an interest in evaluating the default model configuration with the fewest adjustments possible and to utilize the rich regional plot dataset. Testing the performance in a diverse region aids diagnosis of parameter or

35 structural deficiencies. As stated above, the default configuration includes PFT specific physiological variables (no seasonal variation), a constant mortality rate for all forest

PFTs, and a single stem wood allocation equation for all woody PFTs. After evaluation with the default configuration, we tested the use of modified model versions where the physiological parameters and mortality rates were allowed to vary within a PFT by ecoregion (CLM4_eco) and where the stem wood allocation equation in addition to the physiological and mortality rates also varied by ecoregion (CLM4_stem).

For CLM4_eco, parameter values for foliar CN ratios, foliar N content in

Rubisco, leaf longevity, fine root CN ratios, and specific leaf area (SLA) were adjusted according to field plot data from sites in each of the tested ecoregions in the study area.

The PFT physiological variable input file was restructured so that two PFTs were assigned to each ecoregion and surface datasets were modified to reflect the new PFT assignments (Appendix B, Table B1 and B2). The CLM4 mortality module was then modified to assign a different mortality rate based on the PFT. For CLM_stem, the inventory data was used to construct ecoregion-specific allocation to stem wood equations and the equations were added to the CLM4 allocation module (Appendix B,

Table 3 and Figure B1.

Results: Modeled results and regional totals

In general, CLM_stem modeled estimates of stem biomass carbon (Figure 2a) followed a west to east gradient with higher values in the productive mesic western ecoregions (i.e. Coast Range, West Cascades) and lower values in the less productive dry eastern ecoregions (East Cascades, Northern Basin). Stem wood biomass peaks in the

West Cascades at 360 Mg C ha

-1

and is lowest in the Northern Basin at 10 Mg C ha

-1

.

NPP is highest in the Coast Range with values up to 1100 g C m

-2

yr

-1

and lowest in

Northern Basin at less than 100 g C m

-2

yr

-1

(Figure 3.2b). Forest NEP ranges from -200

36 g C m

-2

yr

-1

in the Blue Mountains to 350 g C m

-2

yr

-1

in the Coast Range (Figure 2c) and

R h

ranges from 20 g C m

-2

yr

-1

in the Northern Basin to 750 g C m

-2

yr

-1

in the Coast

Range (Figure 3.2d).

Total simulated regional tree carbon stocks are estimated to be 1.2 Pg C with 61% in the Coast Range and West Cascades alone. For the period from 2001-2006, total forest

NPP and NEP averaged 57.6 and 13.2 Tg C yr

-1

respectively indicating a strong sink in the region. Total modeled harvested carbon equaled 4.9 Tg C yr

-1

and total fire emissions equaled 2.4 Tg C yr

-1

. After accounting for these removals, the forest carbon sink is still positive at 6.2 Tg C yr

-1

.

Results: Spatial evaluation with inventory data

We show ecoregion means of the evergreen needleleaf (ENF) biome because it represents 92% of the forested area in the region and we do not have sufficient regional plot data for the deciduous broadleaf (DBF) biome for evaluation. CLM4 regional modeled mean stem biomass for the period from 2001 – 2006 fell within the observed range of uncertainty in the Coast Range and the West Cascades, was close to observed range of uncertainty in the East Cascades and was underestimated in the Klamath

Mountains (Figure 3a; grey error bars indicate observation uncertainty). However, mean simulated biomass was twice the observed mean in the Blue Mountains. Regional mean

NPP was within the observed range of uncertainty in the Coast Range, East Cascades and

Klamath Mountains, was slightly overestimated in the West Cascades and was again almost twice the observed mean in the Blue Mountains. After modifying the model for ecoregion differences, ecoregion mean stem biomass was within the observed uncertainty in the Blue Mountains, but dropped below the average in the West Cascades. Ecoregion mean NPP was within observation uncertainty in all ecoregions after modification. After modifications, modeled means NPP were not significantly different from observed means

(two-sided p value 0.8 respectively) and modeled means of stem biomass fell within the observed range of uncertainty in the majority of the ecoregions.

Ecoregion means are a good first order approximation of model performance and regional total evaluation, but cannot be used to determine the dynamics that may be

causing any bias and reveal nothing about model fidelity at a smaller scale. Statistical tests of model performance for stem carbon biomass revealed overall better agreement

37 with inventory data using the CLM_stem version of the model. At 12 km spatial resolution, reduced statistics indicated adequate model performance for the region as a whole for both stem biomass and NPP (overall reduced was 3 and 2, respectively).

This was an improvement from the default configuration by 50% for stem biomass and

30% for NPP. Agreement was best in the West Cascades, Blue Mountains, and Coast

Range ( <= 2; Table 3.3) followed by the Klamath Mountains and East Cascades ( <

7). The largest improvements were in the Blue Mountains where was reduced from

15 to 2. Initial CLM4 stem carbon biomass was underestimated by 11 to 41 Mg C ha

-1

in all ecoregions except the Blue Mountains. After modification, stem carbon biomass was consistently underestimated in all ecoregions. Within ecoregions, performance was generally better in older age classes (Table 3.3).

NPP also improved when using CLM4_stem, primarily because overestimation was reduced in the Blue Mountains and West Cascades (Table 3.3). Overestimation in the

Blue Mountains exceeded 100 g C m

-2

yr

-1

with CLM4 before modifications were made.

However, reduced statistics for the Coast, Range, East Cascades and Klamath

Mountains did not significantly improve. NPP was underestimated in the Coast Range and East Cascades by CLM4, improved for the Coast Range with CLM4_stem, but was then overestimated for the East Cascades. As with stem biomass, there was better performance in wet versus dry stands, except in the Klamath Mountains. Performance was better for both metrics in mature stands. This was especially true for the Coast Range and West Cascades where the values for both CLM4 and CLM4_stem were close to 1 for the mature age classes, indicating model-data unity.

Results: Evaluation with tower and supplemental plot data

Simulated monthly values for GPP were compared with 4-5 years of eddycovariance tower data at two sites in the region. At the semi-arid mature ponderosa pine site (Metolius), a reduced value of 0.53 indicated good overall model performance

(values less than 1 indicate good model performance) after accounting for observation

38 uncertainty (Schaefer et al., in press). Observation uncertainty ranged from 12-200% of the observation value for the Metolius site and 13-90% for the Campbell River site.

Maximum GPP is lower for modeled results (Figure 3.4a) and peaks one to two months later than observed values in most of the years compared. Model bias was an underestimation of an average of 27 g C m

-2

yr

-1

for all years compared. Underestimation was highest in 2005, although this was the only year where timing of maximum GPP was equal to observed. There was no noticeable bias in the fall and spring months. At the mesic mature Douglas-fir site (Campbell River), a reduced of 1.1 indicated weaker model performance than the Metolius site, although still nearly within the observed range of uncertainty. Model bias resulted in underestimation of 46 g C m

-2

yr

-1

. Model data mismatch again occurred primarily in summer, with underestimation of summer GPP, and annual GPP up to 200 g C m

-2

yr

-1

. Timing of peak NPP was mostly correct as were fall and spring values.

CLM4 includes direct down-regulation of the photosynthetic rate under nitrogen limitation, which effectively reduces GPP. We also compared GPP before downregulation (Figure 3.4a and Figure 3.4b; “initial GPP”) with tower data and found similar model performance. However, GPP was overestimated by an average of 40 g C m

-2

yr

-1

at the Metolius site, especially in the late summer months. At the Campbell River site

(Figure 3.4b), overall model performance did not improve compared with initial GPP, however summer maximum GPP was much closer to observed. We tested both increasing foliage nitrogen content and altered the equation that calculates biological nitrogen fixation. CLM4 nitrogen fixation is controlled by a single equation that determines the rate as a function of NPP. We increased the rate of fixation for the Campbell River site and found improved simulated summer GPP. However, fall and spring values were also increased resulting in overestimation. The overestimation for fall and spring resulted in correct annual sums of GPP, but no improvement to seasonal model performance.

We compared modeled and observed monthly light use efficiency (LUE) with monthly average air temperature at each tower site (Figure 3.5). We define LUE as the monthly GPP as a function of monthly average incident solar radiation (shortwave

39 radiation). At the Metolius site, observed LUE increased with air temperature to about 10

°C, then declines (Figure 3.5a). In other words, higher GPP is attained at lower radiation levels in the 10°C soils. This pattern was also true for the modeled relationship, but occurred at a higher temperature range, LUE peaks at around 15 °C (Figure 3.5b).

Generally, simulated LUE is lower than observed LUE, and the slope of the relation indicates that simulated LUE is about half that of the observed. At the Campbell River site, there is a smaller increase in LUE with temperature than the Metolius site without the nitrogen availability modifications (Figure 3.5b, grey surface), and there is much less seasonal variation in temperature. As noticed with model performance metrics, introduction of increased N fixation does not increase LUE (the slope), but simply increases the intercept. There is no increase in LUE with temperature in the modeled results. Again, the modeled LUE is about half the observed LUE.

Annual biometric estimates of NPP were available from the Metolius site and are often used as cross-checks with tower data for consistency (Luyssaert et al. 2009).

Modeled annual NPP was underestimated by an average of 24% when compared to biometric estimates of NPP at the Metolius site while modeled GPP was underestimated by an average of 30% compared to the eddy-covariance data (Appendix B, Figure B6).

However, the ration of NPP/GPP was nearly identical for both equaling 37% (observed) and 36% (modeled).

We also compared monthly precipitation with both observed and simulated monthly GPP (Appendix B; Figure B2). For both sites, winter precipitation is much higher and summers are generally dry. Again, GPP is correctly estimated for the higher precipitation months (winter and spring), but underestimated in the low precipitation months (summer).

Results: Model improvement and testing

Finally, we also compared regional plot data of foliar chemistry and morphological characteristics with the physiological file constants used for the modeled runs (Figure 3.6). There is significant variation in both foliar carbon to nitrogen ratios and

40 specific leaf area (SLA, projected) between ecoregions according to plot data. The default configuration for CLM4 allows one foliar CN ratio and one SLA value per PFT.

We experimented with physiological parameter spatial variation (Figure 3.3;

CLM_eco), and found that for the Blue Mountains, overall mean biomass decreased to 71

Mg C ha

-1

and mean NPP decreased to 307 g C m

-2

yr

-1

, both falling within the observed range of uncertainty. The stem carbon reduced statistic decreased from 15 to 2 and the NPP decreased from 5 to 2. Stem carbon bias was reduced from 48 to 9 Mg C ha

-1 and NPP bias decreased from 112 to a slight underestimation of -3 g C m

-2

yr

-1

. We found similar improved relationships in the other ecoregions for NPP with CLM_eco, but stem carbon stocks were subsequently underestimated. CLM4 includes dynamic stem wood allocation, but it is a single equation for all woody PFTs globally. Finally, using

CLM_stem, where we also changed the dynamic stem allocation equation to vary by ecoregion, we found an overall improvement in stem carbon stocks compared to CLM4.

Discussions: Modeled results and regional totals

The regional patterns of carbon stocks, NPP, and NEP are similar to the observed values in this study and prior studies. Stem carbon stocks from over 8000 inventory plots and 200 plots in Oregon showed stem carbon stocks range from 60 – 420 Mg C ha

-1

(Hudiburg et al. 2009, Van Tuyl et al. 2005) in the more mesic ecoregions (Coast Range,

Klamath Mountains, West Cascades). Reported densities from the same studies for the drier ecoregions (East Cascades, Blue Mountains) range from 20 to 100 Mg C ha

-1

. The

CLM4_stem estimates in this study range from 10-400 Mg C in agreement with the reported values. Pacific Northwest temperate forests have some of the highest carbon stocks in the world (Keith et al. 2009), with most of the carbon stored in the tree wood.

Modeled relationships of biomass accumulation over time and with increasing age classes typically show a decrease in tree wood biomass or it reaches an asymptote. However, observations in the Pacific Northwest suggested otherwise (Hudiburg et al. 2009) and modeled allocation to wood biomass in these regions was modified in CLM4 to reflect these regional stand dynamics (CLM4_stem).

41

Modeled NPP ranges from 100-1100 g C m

-2

yr

-1

(CLM4_stem) across the study region agreeing with the reported regional values of 100-900 g C m

-2

yr

-1

(Hudiburg et al.

2009) and falling within the range of 100 to 1600 g C m

-2

yr

-1

reported for temperate and boreal forests (Luyssaert et al. 2008). The regional total NPP for forests is 57.6 Tg C yr

-1

, a value very close to the observed total of 58.2 ± 6.5 Tg C yr

-1

calculated from inventory data (Law et al. 2012). The regional modeled total NEP for 2001-2006 is also close to the reported value from inventory data (13.2 versus 15.2 ± 1.6 Tg C yr

-1

; (Law et al.

2012)), and Biome-BGC modeled results (17.0 ± 10 Tg C yr

-1

; (Turner et al. 2007)), but lower than uncertainty of an inverse modeling estimate (35.4 Tg ± 11 Tg C yr

-1

;

(Göeckede et al. 2010)). The range of NEP reported in a recent study by (Williams et al.

2012) was 0- 270 g C m

-2

yr

-1

for the Pacific Northwest and an overall NEP of 18 Tg C yr

-1

.

Discussion: Spatial evaluation

After modifications were made, stem carbon stocks continued to be underestimated in all ecoregions. Performance was generally better in older age classes, and underestimation of stem biomass was balanced by overestimation of stem biomass in young age classes in some ecoregions. Stem wood allocation is a dynamic function of

NPP rather than age in CLM4. In a modeling assessment of US forest NEP following disturbance, the ecosystem model used (CASA) was altered so that live biomass estimates were constrained by age-accumulation trajectories calculated from FIA data

(Williams et al. 2012). An approach similar to this could be used in CLM4 to remove the age-related bias in stem biomass and would improve on ecoregion-specific allocation we implemented in this study. In the Pacific Northwest, the age-related biomass accumulation trajectories are known to vary by ecoregion and management regime (Van

Tuyl et al. 2005, Hudiburg et al. 2009) with higher potential stem biomass in the more productive ecoregions.

Implementation of ecoregion physiological variation and mortality had a significant impact on NPP performance, with better performance in mature forests for

CLM4_eco and CLM4_stem. This is not surprising considering NPP tends to peak early

in stand development, decline and then remain at a consistent level as stands age. The

42 age-related variation in this response varies by ecoregion and disturbance regime (Law et al. 2004, Hudiburg et al. 2009) and improvement to age-related allocation may help to correct part of the discrepancies. We found underestimation of NPP by CLM4 in the

Coast Range and overestimation in the Blue Mountains was primarily due to the singlevalue parameterization of foliar CN ratios and specific leaf area (Figure 3.6), which are much higher in mesic versus semi-arid forests (Matson et al. 1994). Maximum rates of carboxylation (V cmax

) are determined by SLA and foliar nitrogen content in CLM4. Prior modeling efforts with CLM4 also found significant improvement in GPP estimates by using site-specific parameterization of V cmax

(Bonan et al. 2011.)

NPP continued to be overestimated while stem biomass continued to be underestimated in the West Cascades after implementing CLM4_eco and CL4_stem, although both were improved. Foliar nitrogen content in the Cascade Mountain ecoregion is lower than that of the Coast Range and Klamath Mountain ecoregions and has higher specific leaf area. NPP was reduced with ecoregion-specific parameterization, but this also reduced stem biomass. However, observed West Cascade stem biomass is equal to or higher than Coast Range mean biomass. Stem biomass density is affected by removal rates through fire and harvest. The CLM4 input harvest rates are higher than observed values in the West Cascades (1.9 Tg C yr

-1

vs 1.0 Tg C yr

-1

; Turner et al. 2007)) effectively reducing stem carbon densities over time. Fortunately, these values can be easily adjusted to the correct rates for future simulations. Total regional modeled fire emissions are also overestimated by 1.0 Tg C yr

-1

which could also be contributing to the underestimation bias.

Finally, nitrogen availability has recently been shown to significantly impact the

NPP to GPP ratio, with higher rates of biomass production sites with higher fertility

(Vicca et al., 2012). Our results support this hypothesis, with higher biomass production in the forests with high foliar nitrogen content (Figure 3.2 and Figure 3.6). Thus, the model represents nitrogen influences on allocation and productivity quite well.

Discussion: Seasonal evaluation

Simulating GPP correctly is important because most of the subsequently calculated carbon fluxes are calculated from GPP. In a model-data synthesis activity including 39 flux sites and 26 models (CLM4 was not included) GPP was found to be overestimated in the spring and fall and underestimated in the summer resulting in poor

43 model performance at most flux sites (Schaefer et al, in press). We found CLM4 simulated GPP at the Metolius site to be close to the observed range of uncertainty

(reduced = 0.60) and GPP was not over-predicted in fall and spring resulting in overall fair model performance. However, GPP was under-predicted in the summer months (Figure 3.6a), and it peaked one month later than observed values. LUE was determined as the primary driver of underestimation of GPP in summer months in the model-data synthesis. Our analysis agrees with this as we found LUE to be on average less efficient in the modeled results than the observations. However, this does not necessarily mean the low GPP bias in the model is caused by incorrect representation of

LUE. We also found a different range of temperature thresholds for the range of LUE values indicating PFT-specific temperature thresholds may vary, especially in the drier more extreme climates found in the eastern portion of our study region. GPP was also found to be underestimated for months with little to no precipitation (i.e. summer) indicating soil water availability or plant water use efficiency need further investigation.

We also noted the “initial GPP” before down-regulation as a result of nitrogen limitation had a positive summer bias, and while down-regulation brought the majority of the simulated GPP monthly values closer to observed, there was too much constraint in the summer months.

At the Campbell River site, model performance was similar with the same tendency to underestimate summer GPP. Foliar nitrogen content and availability are both higher at the Campbell River site than the Metolius site. However, since LUE is underestimated at both sites, V cmax

must also be underestimated. V cmax

increases with increasing nitrogen availability and fraction of leaf nitrogen in Rubisco which could be calculated from plot data collected in the region to improve parameterization of leaf nitrogen content and Rubisco activity in the Pacific Northwest.

44

Finally, the tower site simulations were forced with the regional downscaled data for the Metolius site and the NCEP data for the Campbell River site until 1950. While this is typical for point simulations where multiple years of meteorological forcing data are unavailable, there are recent modeling activities where site data are being used for the entire simulation period and initial model results tend to achieve overall better model performance with site data for the duration of the simulation period (P. Thornton, personal communication)

Summary of results and recommendations for model improvement

CLM4 was evaluated against inventory data in PNW forests using the default configuration and regional parameterization. Initial results indicated overall underestimation of stem biomass except in the semi-arid Blue Mountains where it was overestimated by 48 Mg C m

-2

. There was good general agreement with observed NPP values. However, modeled ecoregion mean NPP was overestimated in the Blue

Mountains and mesic West Cascades. Following initial default parameterization, model improvements were made to account for ecoregion differences in the physiological variables, foliar N content and mortality, and this resulted in an overall improvement in

NPP estimates (all ecoregions fell within the observed range of uncertainty). Changing the stem wood allocation algorithm further improved the results, however wood carbon stocks were still underestimated in the West Cascades and Klamath Mountains

(CLM4_stem). Adjusting stem wood allocation changes with age should improve results.

The amount of wood mass harvested was overestimated in the West Cascades compared to observed values which could be causing the reduced modeled stem carbon stocks compared to observations. Harvest values are easily corrected or changed. Fire emissions are a function of the percentage of wood combusted and the burn area. The fire module can be easily modified to reduce the amount of stem biomass combusted during wildfire to match historical rates. The percentage of area burned is a more complicated prediction, a process that was improved by Kloster et al. (2011) but not yet incorporated into public release versions of CLM4.

At 12 km spatial resolution, reduced statistics indicated adequate model performance in all ecoregions for both stem biomass and NPP (reduced was equal 3 and 2, respectively). This was an improvement from the default configuration by 50% for

45 stem biomass and 30% for NPP. Within ecoregions, there was good performance ( < 2) in the Coast Range and West Cascades. There was generally better performance in mature stands and apart from the Klamath Mountains, better performance in wet stands.

Evaluation of CLM4 monthly GPP with eddy-covariance tower data revealed good model-data agreement from October to May at the mesic site (ca1; Figure 3.4b) and

August – April at the semi-arid site (me2; Figure 3.4a). Summer GPP was underestimated at both sites due to several possible reasons including variation in seasonal V cmax response to temperature and nitrogen dynamics. We found a different range of temperature thresholds for the range of LUE values indicating PFT-specific temperature thresholds may vary, especially in the drier more extreme climates found in the eastern portion of our study region. Adjustments to V cmax

by increasing nitrogen availability improved summer GPP estimates, but reduced model fidelity for the remainder of the months suggesting nitrogen availability alone does not improve seasonality. Improvement to V cmax

seasonality in CLM4 will require adjustment to other factors such as response to temperature and soil water in order to improve summer GPP.

References

Bonan, G. B., P. J. Lawrence, K. W. Oleson, S. Levis, M. Jung, M. Reichstein, D. M.

Lawrence, and S. C. Swenson. 2011. Improving canopy processes in the

Community Land Model version 4 (CLM4) using global flux fields empirically inferred from FLUXNET data. J. Geophys. Res. 116:G02014.

Bonan, G. B., K. W. Oleson, R. A. Fisher, G. Lasslop, and M. Reichstein. 2012.

Reconciling leaf physiological traits and canopy flux data: Use of the TRY and

FLUXNET databases in the Community Land Model version 4. J. Geophys. Res.

117:G02026.

Edburg, S. L., J. A. Hicke, D. M. Lawrence, and P. E. Thornton. 2011. Simulating coupled carbon and nitrogen dynamics following mountain pine beetle outbreaks in the western United States. J. Geophys. Res. 116:G04033.

Farquhar, G. D., S. v. Caemmerer, and J. Berry. 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 149:78-90.

Göckede, M., A. M. Michalak, D. Vickers, D. P. Turner, and B. E. Law. 2010.

Atmospheric inverse modeling to constrain regional-scale CO2 budgets at high spatial and temporal resolution. J. Geophys. Res. 115:D15113.

Groenendijk, M., A. J. Dolman, C. Ammann, A. Arneth, A. Cescatti, D. Dragoni, J. H. C.

Gash, D. Gianelle, B. Gioli, G. Kiely, A. Knohl, B. E. Law, M. Lund, B.

Marcolla, M. K. van der Molen, L. Montagnani, E. Moors, A. D. Richardson, O.

Roupsard, H. Verbeeck, and G. Wohlfahrt. 2012a. Seasonal variation of photosynthetic model parameters and leaf area index from global Fluxnet eddy covariance data. J. Geophys. Res. 116:G04027.

46

Groenendijk, M., A. J. Dolman, M. K. van der Molen, R. Leuning, A. Arneth, N.

Delpierre, J. H. C. Gash, A. Lindroth, A. D. Richardson, H. Verbeeck, and G.

Wohlfahrt. 2012b. Assessing parameter variability in a photosynthesis model within and between plant functional types using global Fluxnet eddy covariance data. Agricultural and Forest Meteorology 151:22-38.

Harmon, M. E., W. K. Ferrell, and J. F. Franklin. 1990. Effects on carbon storage of conversion of old-growth forests to young forests. Science 247:699-702.

Hicke, J. A., J. C. Jenkins, D. S. Ojima, and M. Ducey. 2008. Spatial patterns of forest characteristics in the western United States derived from inventories. Ecological

Applications.

Hudiburg, T., B. E. Law, D. P. Turner, J. Campbell, D. Donato, and M. Duane. 2009.

Carbon dynamics of Oregon and Northern California forests and potential landbased carbon storage. Ecological Applications 19:163-180.

Hudiburg, T. W., B. E. Law, C. Wirth, and S. Luyssaert. 2011. Regional carbon dioxide implications of forest bioenergy production. Nature Clim. Change 1:419-423.

Hurtt, G. C., S. Frolking, M. G. Fearon, B. Moore, E. Shevliakova, S. Malyshev, S. W.

Pacala, and R. A. Houghton. 2006. The underpinnings of land-use history: three centuries of global gridded land-use transitions, wood-harvest activity, and resulting secondary lands. Global Change Biology 12:1208-1229.

IPCC. 2007. IPCC expert meeting report: Technical Summary, Towards new scenarios for analysis of emissions, climate change, impacts, and response strategies.

Intergovernmental Panel on Climate Change, Cambridge.

Jenkins, J. C., D. C. Chojnacky, L. S. Heath, and R. A. Birdsey. 2003. National-Scale

Biomass Estimators for United States Tree Species. Forest Science 49:12-35.

Keith, H., B. G. Mackey, and D. B. Lindenmayer. 2009. Re-evaluation of forest biomass carbon stocks and lessons from the world's most carbon-dense forests.

Proceedings of the National Academy of Sciences, USA 106:11635-11640.

47

Krishnan, P., T. A. Black, R. S. Jassal, B. Chen, and Z. Nesic. 2009. Interannual variability of the carbon balance of three different-aged Douglas-fir stands in the

Pacific Northwest. J. Geophys. Res. 114:G04011.

Law, B., T. Hudiburg, and S. Luyssaert. 2012. Thinning Effects on Forest Productivity:

Consequences of Preserving Old Forests and Mitigating Impacts of Fire and Drought.

Plant Ecology & Diversity doi: 10.1080/17550874.2012.679013

Law, B. E., O. J. Sun, J. Campbell, S. Van Tuyl, and P. E. Thornton. 2003. Changes in carbon storage and fluxes in a chronosequence of ponderosa pine. Global Change

Biology 9:510-524.

Law, B. E., D. Turner, J. Campbell, O. J. Sun, S. Van Tuyl, W. D. Ritts, and W. B.

Cohen. 2004. Disturbance and climate effects on carbon stocks and fluxes across

Western Oregon USA. Global Change Biology 10:1429-1444.

Lawrence, D., K. W. Oleson, M. G. Flanner, P. E. Thorton, S. C. Swenson, P. J.

Lawrence, X. Zeng, Z.-L. Yang, S. Levis, K. Skaguchi, G. B. Bonan, and A. G.

Slater. 2011. Parameterization Improvements and Functional and Structural

Advances in Version 4 of the Community Land Model. J. Adv. Model. Earth Syst.

3:27.

Luyssaert, S., E. D. Schulze, A. Borner, A. Knohl, D. Hessenmoller, B. E. Law, P. Ciais, and J. Grace. 2008. Old-growth forests as global carbon sinks. Nature 455:213-

215.

Matson, P., L. Johnson, C. Billow, J. Miller, and R. Pu. 1994. Seasonal Patterns and

Remote Spectral Estimation of Canopy Chemistry Across the Oregon Transect.

Ecological Applications 4:280-298.

Moss, R. H., J. A. Edmonds, K. A. Hibbard, M. R. Manning, S. K. Rose, D. P. van

Vuuren, T. R. Carter, S. Emori, M. Kainuma, T. Kram, G. A. Meehl, J. F. B.

Mitchell, N. Nakicenovic, K. Riahi, S. J. Smith, R. J. Stouffer, A. M. Thomson, J.

P. Weyant, and T. J. Wilbanks. 2010. The next generation of scenarios for climate change research and assessment. Nature 463:747-756.

48

NRC. 2010a. Advancing the Science of Climate Change. National Academy of Sciences,

National Academies Press, Washington D.C.

NRC. 2010b. Verifying Greenhouse Gas Emissions National Academies Press,

Washington, D.C.

Oleson, K. W., D. M. Lawrence, G. B. Bonan, M. G. Flanner, E. Kluzek, P. J. Lawrence,

S. Levis, S. C. Swenson, P. E.Thornton, A. Dai, M. Decker, R. DIckinson, J.

Feddema, C. L. Heald, F. Hoffman, J. F. Lamarque, N. Mahowald, G.-Y. Niu, T.

Qian, J. Randerson, S. Running, K.Sakaguchi, A. Slater, R. Stockli, A. Wang, Z.-

L. Yang, X. Zeng, and X. Zeng. 2010. Technical description of version 4.0 of the

Community Land Model. NCAR Tech. Note NCAR/TN-478+STR, 257.

Omernik, J.M. 1987. Ecoregions of the conterminous United States. Map (scale

1:7,500,000). Annals of the Association of American Geographers 77(1):118-125.

Pacala, S. W. 2010. Verifying Greenhouse Gas Emissions: Methods to Support

International Climate Agreements. NRC The National Academies Press,

Washington, D.C.

Qian, T., A. Dai, K. E. Trenberth, and K. W. Oleson. 2006. Simulation of Global Land

Surface Conditions from 1948 to 2004. Part I: Forcing Data and Evaluations.

Journal of Hydrometeorology 7:953-975.

Salathé, E., L. Leung, Y. Qian, and Y. Zhang. 2010. Regional climate model projections for the State of Washington. Climatic Change 102:51-75.

Salathe, E. P., P. W. Mote, and M. W. Wiley. 2007. Review of scenario selection and downscaling methods for the assessment of climate change impacts on hydrology in the United States pacific northwest. International Journal of Climatology

27:1611-1621.

Schaefer, K. M., C. R. Schwalm, C. A. Williams, M. Arain, A. G. Barr, J. M. Chen, and

K. J. Davis. In Press. A model-data comparison of Gross Primary Productivity:

Results from the North American Carbon Program site synthesis. Journal of

Geophysical Research.

Schwalm, C. R., C. A. Williams, K. Schaefer, R. Anderson, M. A. Arain, I. Baker, A.

Barr, T. A. Black, G. Chen, J. M. Chen, P. Ciais, K. J. Davis, A. Desai, M. Dietze,

D. Dragoni, M. L. Fischer, L. B. Flanagan, R. Grant, L. Gu, D. Hollinger, R. C.

Izaurralde, C. Kucharik, P. Lafleur, B. E. Law, L. Li, Z. Li, S. Liu, E. Lokupitiya,

Y. Luo, S. Ma, H. Margolis, R. Matamala, H. McCaughey, R. K. Monson, W. C.

Oechel, C. Peng, B. Poulter, D. T. Price, D. M. Riciutto, W. Riley, A. K. Sahoo,

M. Sprintsin, J. Sun, H. Tian, C. Tonitto, H. Verbeeck, and S. B. Verma. 2010. A

model-data intercomparison of CO2 exchange across North America: Results from the North American Carbon Program site synthesis. J. Geophys. Res.

115:G00H05.

Smithwick, E. A. H., M. E. Harmon, S. M. Remillard, S. A. Acker, and J. F. Franklin.

2002. Potential upper bounds of carbon stores in the Pacific Northwest.

Ecological Applications 12:1303-1317.

Thomas, C. K., B. E. Law, J. Irvine, J. G. Martin, J. C. Pettijohn, and K. J. Davis. 2009.

49

Seasonal hydrology explains interannual and seasonal variation in carbon and water exchange in a semiarid mature ponderosa pine forest in central Oregon. J.

Geophys. Res. 114:G04006.

Thornton, P. E., B. E. Law, H. L. Gholz, K. L. Clark, E. Falge, D. S. Ellsworth, A. H.

Goldstein, R. K. Monson, D. Hollinger, and M. Falk. 2002. Modeling and measuring the effects of disturbance history and climate on carbon and water budgets in evergreen needleleaf forests. Agricultural and Forest Meteorology

113:185-222.

Thornton, P. E., S. W. Running, and M. A. White. 1997. Generating surfaces of daily meteorological variables over large regions of complex terrain. J. Hydrol.

190:214–251.

Thornton, P. E., and N. E. Zimmermann. 2007. An improved canopy integration scheme for a land surface model with prognostic canopy structure. Journal of Cllimate

20:3902-3923.

Turner, D. P., D. Ritts, B. E. Law, W. Cohen, Z. Yang, T. Hudiburg, J. L. Campbell, and

M. Duane. 2007. Scaling net ecosystem production and net biome production over a heterogeneous region in the western United States. Biogeosciences 4:597-612.

Van Tuyl, S., B. E. Law, D. P. Turner, and A. I. Gitelman. 2005. Variability in net primary production and carbon storage in biomass across Oregon forests--an assessment integrating data from forest inventories, intensive sites, and remote sensing. Forest Ecology and Management 209:273-291.

White, M. A., P. E. Thornton, S. W. Running, and R. R. Nemani. 2000. Parameterization and Sensitivity Analysis of the BIOMEBGC Terrestrial Ecosystem Model: Net

Primary Production Controls. Earth Interactions 4:1-85.

Williams, C. A., G. J. Collatz, J. Masek, and S. N. Goward. 2012. Carbon consequences of forest disturbance and recovery across the conterminous United States. Global

Biogeochem. Cycles 26:GB1005.

Wirth, C., K. Ogle, A. Nuske, J. T. Weedon, W. K. Cornwell, and J. Cornelissen. 2010.

50

Global analysis of wood decomposition – the roles of environment, taxonomy, and traits. in. Max Planc Institute for Biogeochemistry.

Table 3.1: Ecoregion and tower site characteristics in study region

Ecoregion Forest (ha)

(% of

Total)

Coast Range (CR) 2043332

(17)

Number plots

Private /

Public

294 / 216

Stand Age

Private /

Public

34 / 75

West Cascades (WC)

Klamath Mountains

(KM)

Willamette Valley

(WV)

East Cascades (EC)

Blue Mountains

(BM)

Columbia Plateau

(CP)

Northern Basin (NB) 253690

Metolius Mature Site

(me2)

Campbell River

Mature Site (ca1)

2693263

(22)

1302111

(10)

501793

(4)

2099866

(17)

3364151

(27)

88922

(<1)

(2)

--

--

168 / 512

121/192

93/17

155 / 395

243 / 614

80 / 47

80 / 130

1

1

50 / 140

59 / 106

43 / 61

64 / 94

71 /100

80 / 47

80 / 130

80

56

Dominant Forest Types MAP

(mm yr

Douglas-fir, Sitka Spruce,

Redwood, Western Red

Cedar, Fir-hemlock

Douglas-fir, Hemlock,

Mixed Conifer, Red Fir,

Western Red Cedar

Mixed Conifer, Mixed

Evergreen, Red Fir,

Douglas-fir, Riparian, Oak

Douglas-fir, Hemlock,

Riparian

Ponderosa Pine, Mixed

Conifer, Juniper, Pine,

Red Fir

Mixed Conifer, Ponderosa

Pine, Juniper, Spruce-Fir

Mixed Conifer, Ponderosa

Pine, Riparian

Juniper, Aspen, Pinyon-

Juniper, Ponderosa Pine,

Mountain Mahogany

Ponderosa Pine (secondary growth, privately owned)

Douglas Fir (secondary growth, privately owned)

1742

1688

1549

1280

630

552

330

304

434

1256

-1

)

MAT

(C°)

11.0

8.8

11.5

11.0

9.1

7.3

9.7

9.7

7.6

8.7

Table 3.2: Model simulations. Transient CO

2

, nitrogen deposition (Ndep), and land cover files are annual files covering the period from 1850-2006. Ndep and land cover are spatially interpolated using the CLM toolkit from a half-degree global dataset while the

CO

2

file is a single value for the entire region which changes annually. Simulations marked with and ‘*’ indicate tower/point simulations (ca1 = Campbell River and me2 = Metolius Mature Site). Downscaled regional data was unavailable for the ca1 site.

Name

S_Oregon

S_ca1

S_me2

*

*

C_Default

C_eco

C_eco_stem

Years

650

650

650

Climate Forcing

(Years)

PNW (1900-1949)

NCEP (1948-1972)

PNW (1900-1949)

1850-2006 PNW (1900-1949)

1850-2006 PNW (1900-1949)

CO

2

/

Ndep

Spinup

1850

1850

1850

Land

Cover

1850

1850

1850

Control and Calibration

1850 1850

2000 2000

1850-2006 PNW(1900-1949) 2000 2000

Mortality

Rate

2 %

1%

1%

1 % varied varied

Physiology Model

PNW

Ca1

Me2

PNW

PNW_eco

CLM4

CLM4

CLM4

CLM4

CLM4

_eco

PNW_eco CLM4

_stem

T_PNW_default

T_PNW_ eco

Transient and Modified

1850-2006 PNW (1900-1949;1900-2006) transient transient 1%

1850-2006 PNW (1900-1949;1900-2006) transient transient varied

T_PNW_eco_stem

T_ca1

T_me2

*

*

T_ca1_nitrogen

*2

1

1850-2006 PNW (1900-1949;1900-2006)

1850-2010 NCEP (1948-1972); Tower

1850-2010 PNW (1900-1949; Tower

1850-2010 NCEP (1948-1972); Tower transient transient varied transient transient 1.0 % transient transient 1.0 % transient transient 1.0 %

PNW CLM4

PNW_eco CLM4

_eco

PNW_eco CLM4

Ca1

Me2

Ca1

_stem

CLM4

CLM4

CLM4

_nitr

1

The stem wood to leaf allocation equation was modified to change with ecoregion (location).

2

The nitrogen dynamics were modified to include increased biological fixation.

50

Table 3.3. Stem carbon and NPP reduced precipitation regime a

, and age group b statistic and model bias by ecoregion, annual

. Values marked with an ‘*’ indicate good model performance. ( . Reduced values close to 1 are within observation uncertainty.

Stem Carbon: statistics

Eco- region

CR

WC

KM

EC

BM

5

17

Overall 12

6

6

4

CLM4 simulations

Precipitation Age Group

-

Dry Wet Y

3 4

3

1

*

2

8

14

2

15

70

18

2

8

5

17

8

2

7

4

15

7

2

1

*

5

3

4

CLM4_stem simulations

Precipitation Age Group

1

*

M Total Dry Wet Y M

2 - 3 4 1

*

<2

6

5

2

3

8 <2

*

1

*

14

14

7

6

5

3

3

Total

2

<2

*

6

4

2

3

NPP: statistics

Ecoregion

Precipitation

CLM4 simulations

Age Group

CR

WC

KM

EC

BM 5

Overall 4

-

Dry Wet Y

3

1

*

2

1

*

1

*

7

<2

*

2

8

6

8

3

2

18

7

1

*

1

*

7

<2

*

5

5

1

1

6

2

5

3

*

*

-

4

1

*

3

2

3

1

CLM4_stem simulations

Precipitation

<2

7

*

<2

1

*

2

*

*

Age Group

M Total Dry Wet Y M

3

6

14

8

6

1

6 1

6

3

2

4

*

*

Total

<2

*

1

*

5

3

2

2

Stem Carbon: Bias

Ecoregion

Precipitation

Dry Wet

CLM4 simulations

Age Group

Y

CLM4_stem simulations

Precipitation Age Group

M Total Dry Wet Y M Total

CR

WC

KM

EC

BM 54

Overall 28

-

13

-85

-4

-19 32

-11 74

-40 3

-49 20

2 77

-17 46

-100 -19

-40 -9

-53

-17

-41

-11

45

-16

48

4

-

-7

-77

-2

-4

-5

-17 36 -100 -17

-37 45 -65 -36

-51 -7 -63

-48 18 -15

-50

-9

-36 21 -11

-32 29 -41

-7

-18

Ecoregion

CR

WC

KM

EC

BM

CLM4 simulations

Precipitation Age Group

CLM4_stem simulations

Precipitation Age Group

-

Dry Wet Y

-88 -88

M

-90

Total

-88

Dry Wet

13

Y

20

M

-2

93 94 79 97 96 78 55 24 62

Total

13

58

-221 105 22

-30 -132 17

92

-54

85

-47

-140 93

49 -60

-4

87

85

30

81

35

133 -66 149 104

NPP: Bias

112

Overall 69 14 19 50 43 a

Dry = < 800 mm/yr, Wet = >800 mm/yr b

Young = < 60 years, Mature = > 60years

17

27

-128 40 -6

25 32 29

2

31

Figure 3.1. Study region area (Oregon) divided by ecoregions (BM = Blue Mountains,

CR= Coast Range, CP = Columbia Plateau, EC = East Cascades, KM = Klamath

Mountains, NB = Northern Basin, SR = Snake River, WC = West Cascades, WV =

Willamette Valley). Green circles represent FIA plots used for evaluating annual model output, blue triangles are the supplemental plots used for parameterization, and red squares are the AmeriFlux tower sites used for seasonal validation (Campbell River site not shown).

51

52

Figure 3.2: Maps of modeled CLM4_stem estimates of a) Stem Biomass, b) NPP, c)

NEP, and d) heterotrophic respiration (Rh). Biomass is in units of Mg C ha

-1 others are in units of g C m

-2

yr and all

-1 . Estimates are from the ‘all transient’ case which is representative of actual historical conditions and includes changing climate, CO

2

and nitrogen deposition levels, and land use and land cover change.

A

53

Figure 3.3: Ecoregion means of modeled and observed evergreen needleleaf (ENF) PFT values for the period from 2001-2006. The ENF biome covers 92% of the forested area in the study region. Bar plots show ecoregion ENF means and the associated observation uncertainty. a) Stem Biomass, and b) NPP. Estimates are from the ‘all transient’ case which is representative of actual historical conditions and includes changing climate, CO

2 and Nitrogen levels, and land use and land cover change.

B

54

Figure 3.4: Simulated monthly GPP versus observed GPP at two FLUNET eddycovariance tower sites in the study region. Solid circles represent tower observations, hollow squares are modeled values of GPP before downregulation due to nitrogen limitation, and thick-lined crosses are GPP after downregulation. Grey bars represent observed estimate uncertainty. (A) Metolius tower site in Oregon, USA. (B) Campbell

River tower site, British Columbia, Canada. The thin lined crosses represent the model experiment where nitrogen availability was increased.

55

Figure 3.5: 3D surface plots of shortwave radiation versus GPP (LUE) versus monthly average temperature. (A). Observed LUE versus temperature at the Metolius site, (B)

Modeled LUE versus temperature at the Metolius site, (C) Observed LUE versus temperature at the Campbell River site, and (D) Modeled LUE versus temperature at the

Campbell River site for the CLM4 and the modified nitrogen availability versions of the model code.

56

Figure 3.6: Observed values for ENF leaf carbon to nitrogen ratios (A) and specific leaf area (B) from over 100 supplemental plots in the region. Black bar represents the value used for single-value per PFT parameterizations versus the grey bars that were used for

CLM_eco and CLM_stem simulations.

57

Chapter 4: Long-term effects of bioenergy harvest on the carbon balance of Pacific

Northwest forests under changing climate and climate-related disturbance

Abstract

Bioenergy is quickly becoming an important fossil fuel replacement source because it can provide energy security, economic development, and is expected to reduce greenhouse gas emissions compared to fossil fuel use. There is continued debate about actual reductions, but recent studies indicate potential areas where CO

2

reductions may be possible where the land based carbon sink has been reduced due to increased mortality from fire, insects, and drought. Here, we examine forest carbon response and life cycle assessment (LCA) of net carbon emissions following varying combinations of bioenergy management scenarios in Pacific Northwest forests for the period from 2010-2100. We use the NCAR CLM4 model combined with a regional atmospheric forcing dataset and account for future environmental change using the moderate IPCC RCP4.5 scenario. At the end of the 21 st

century, predicted regional net ecosystem production (NEP) is 13.7 Tg

C yr

-1

(107 g C m

-2

yr

-1

) for business-as-usual (BAU) conditions compared with the current NEP of 13.2 ± 1.6 Tg (103 g C m

-2 yr

-1

). There is no significant influence on NEP by changing climate, nitrogen deposition, and increasing CO

2

concentrations in the long term for the moderate RCP4.5 scenario where CO

2

rises to 550 ppm by 2100. Increases in

BAU net primary production (NPP) are accompanied by increases in heterotrophic respiration (R h

) from a warmer climate, resulting in no change in NEP. BAU net biome production (NBP; NEP – harvest removals and fire emissions) averages 38 ± 8 g C m

-2 yr

-1

, a decline compared to the current NBP for the study region of 61 ± 14 g C m

-2

yr

-1

(Hudiburg et al. 2011). Nitrogen deposition, precipitation, and harvest rates do not change significantly and therefore changes in predicted carbon fluxes are due to climate warming, CO

2

fertilization, and fire. Bioenergy management strategies include a repeated thinning harvest where ~35% of stand biomass was removed in fire prone areas with high stand density or productive areas with high stand density. They also include a repeated clearcut harvest where ~85% of stand biomass was removed in highly productive areas and a single salvage harvest where ~75% of biomass was removed in areas with projected

insect-related mortality. None of the bioenergy management scenarios reduce net emissions to the atmosphere compared to BAU by the end of the 21 st

century. Forest regrowth and reduced fire emissions are not large enough to balance the wood removals from harvest. Moreover, the substitution of wood for fossil fuel energy and .products is

58 not large enough to offset the wood losses through decomposition and combustion.

However, in some ecoregions (Blue Mountains and East Cascades), emissions from the thinning harvests are beginning to improve over BAU at the end of the century and could lead to net reductions over a longer time period (> 100 years). For salvage logging, there is no change compared to BAU emissions by the end of the 21 st

century because the treatment area is minimal compared to the other treatments and only performed once.

Introduction

Wood bioenergy in US Pacific Northwest forests is currently supplied by forest thinnings intended to improve forest health and by forest residues from current harvest. In efforts to improve forest health, create jobs, restore habitat, provide energy security, and reduce greenhouse gas emissions, new management plans have been considered or implemented to increase utilization of forest biomass for wood products and bioenergy.

Recently, there has been a growing interest in integrating wood bioenergy with rotation-based wood product harvest on industrial forestland, similar to some European countries (Routa et al. 2011). Unlike traditional harvest and thinning practices where the non-merchantable wood is slowly decomposed or burned on site, rotation-based harvest removes all of the woody biomass offsite including the residues. Consequently, the natural transfer of live wood to the woody detritus, litter, and eventually the soil carbon and nitrogen pools through mortality will be lessened, decreasing the baseline land-based carbon stocks (Mitchell et al. 2012, Schulze et al. 2012). In turn, nitrogen losses due to detritus removals may limit future carbon uptake potential. Policy and management plans based on current forest condition will influence the extent to which the forest industry utilizes harvest residues, forest thinnings, and rotation-based harvest to meet energy demands for replacement of fossil fuels intended to reduce greenhouse gas (GHG) emissions to the atmosphere.

The basic assumptions, justifying an increased harvest to supply bioenergy, are

59 that forest bioenergy will be carbon neutral, with no additional CO

2

emissions because carbon in wood was recently removed from the atmosphere. It is also assumed to be an effective means for reducing greenhouse gas emissions because the forest will regrow new biomass to replace removed carbon and that bioenergy emits less carbon per unit of energy than fossil fuel-derived energy (Lippke et al. 2011, Malmsheimer et al. 2011).

This is contrary to comprehensive studies that account for the baseline carbon sink and show an increase in emissions compared to business-as-usual management in the next several decades. The ‘Manumet’ study in Massachusetts found that utilizing forest biomass for bioenergy production did not provide enough carbon substitution benefits to replace the reduced forest carbon storage (Walker et al. 2010). Another study in Ontario, showed CO

2

increases for the next 16-38 years for combined heat and power (CHP) use and hundreds of year for ethanol conversion (McKechnie et al. 2011). In addition to the amount of time required for bioenergy substitutions to repay the forest carbon lost due to harvest, it can take hundreds of years for bioenergy production to substitute the amount of carbon that would be stored if the forest were left unharvested (Mitchell et al. 2012).

However, these studies did not examine the expected changes in the baseline carbon sink owing to climate change, which for the western US is predicted and observed to affect species composition (Coops and Waring 2011), lead to increased mortality from drought stress (van Mantgem et al. 2009, Breshears et al. 2008, van Mantgem et al. 2009,

McDowell et al. 2010), insect outbreaks (Kurz et al. 2008, Raffa et al. 2008), and fire

(Rogers et al. 2011). Forest carbon sinks will change as climate and societal demands change. The change ultimately depends on the interplay of regional conditions such as current sink strength, climate, fire regime, absence of catastrophic disturbances, etc.

(Hudiburg et al. 2011, Hudiburg et al. 2009, Zhang et al. 2012).

Here, we go beyond recent studies by simulating the effects of varying bioenergy harvest scenarios on forest carbon and nitrogen dynamics and by accounting for future environmental change (climate, increasing atmospheric CO

2

and nitrogen deposition).

We focus on the forests of Oregon, USA, a state characterized by a wide range of climate and productivity. We provide potential bioenergy management solutions that account for

60 carbon and subsequent global change mitigation consequences. The specific questions we answer are:

1. What are the interactive effects of changing climate, increasing atmospheric

CO

2

, N deposition and land use change on net ecosystem production (NEP)?

2. How do the varying management scenarios affect carbon stocks and fluxes at different spatial scales?

3. Does bioenergy management increase or decrease net CO

2

emissions to the atmosphere compared to BAU by the end of the 21 st

century?

4. For which ecoregions do net CO

2

emissions decrease compared to BAU for bioenergy management?

Materials and Methods

Ecosystem models provide a means to analyze the multiple interactions of environmental change with management, simulate the effects separately and together, and evaluate the relative importance of each factor. We modeled the interactive effects of environmental change and bioenergy management using a modified version of the NCAR

CLM4 model developed for use in the study region (Hudiburg et al., in prep; dissertation

Chapter 3) combined with a regional atmospheric forcing dataset. The output of the

NCAR CLM4 simulations contain harvested biomass that was used in an off-line LCA model to quantify the net CO

2

emissions of the bioenergy scenarios.

Model description and parameterization

The Intergovernmental Panel for Climate Change (IPCC) synthesizes estimates of future climate change impacts on terrestrial carbon cycling through the use of a specific set of nationally recognized global circulation models. Among them is the Community

Earth System Model (CESM) of the National Center for Atmospheric Research (NCAR).

The land model component (Community Land Model; CLM4) has been used to evaluate and predict the net carbon uptake and loss from terrestrial biomes, particularly forests

(Thornton and Zimmermann 2007, Bonan et al. 2011) and has been scientifically

61 validated with regional observations (Ch.3). We use a modified version of CLM4 described by Hudiburg et al., (in prep; Ch. 3) that accounts for within PFT and ecoregion variation in physiological parameters, mortality, and stem wood allocation. Regional evaluation with this model version found good model agreement in all ecoregions with historical annual net primary production data and slight underestimation of woody biomass in the highly productive ecoregions. Historical regional annual totals of NPP,

NEP, and NBP were also within or very close to the observed and/or modeled range of uncertainty from inventory, tower, process model, and atmospheric inversion model estimates in the region suggesting the model should be able to make reasonable predictions about the effects of future land use change on forest carbon and nitrogen fluxes.

Through the combined use of regional climate data, physiological characteristics, and site physical attributes and history, CLM4 estimates the daily carbon, nitrogen, and water fluxes between the atmosphere and the vegetation and soils. Active model components used in this study include biogeophysics, hydrology, and biogeochemistry.

The land surface is divided into grid cells and the vegetated portion of each grid cell is further divided into plant functional types (PFTs) designated by a percentage cover of the grid cell. Each PFT has its own leaf and stem area index and canopy height and each PFT competes for water and nutrients on a single soil column.

Recent model improvements include an integrated transient land cover and land use change dataset that includes wood harvest (Lawrence et al. 2011). The transient land use dataset has been formatted for use by CLM4 from a global historical transient land use and land cover change dataset (Hurtt et al. 2006) covering the period from 1850–

2005. The dataset is not a dynamic vegetation model where PFT changes occur based on climate, stress, and disturbance drivers overtime. This dataset is a prescribed humaninduced disturbance history of forests which we can use to test the model’s ability to simulate carbon dynamics following clearcut or thinning harvests. We use this embedded dynamic PFT framework to implement our future land use change scenarios (described below). We use the dataset to prescribe the annual fraction of land that is transformed

62 from one PFT to another and to prescribe the harvest rates for each PFT within a grid cell over the simulation period.

CLM4 includes PFT-specific physiological variables, a constant mortality rate for all forest PFTs, and a single stem wood allocation equation for all woody PFTs. After we evaluated historical simulations with eddy-covariance tower and forest inventory data, we modified CLM4 to enable the physiological parameters and mortality rates to vary within a PFT by ecoregion and to allow the stem wood allocation equation to vary by ecoregion

(Ch.3). We use the modified version for all simulations in this study. The modifications and prescribed parameters are provided in Appendix B.

Model simulations of management scenarios

We modeled future carbon and nitrogen dynamics from 2010-2100 for the state of

Oregon located in the Pacific Northwest region of the United States (Figure 4.1). Model spin-up, regional and plant functional type calibration, historical simulations and model development were completed for a separate evaluation analysis of the region. After modifications were made to CLM4 to more closely represent the ecoregion-specific variability of a single plant functional type’s physiology, mortality, and allocation patterns, we found good model agreement for spatial patterns of NPP and stem biomass with inventory data (Ch 3). Modeled ecoregion means of NPP were not significantly different from observed ecoregion means (two-sided p value = 0.8) and modeled means of stem biomass fell within the observed range of uncertainty for the majority of the ecoregions with better performance in mesic ecoregions and older age classes. There was also overall good model agreement with seasonal GPP at eddy-covariance tower locations in the region (reduced statistics were close to 1, Ch 3). There was some underestimation of summer GPP (25-100 g C m

-2

month

-1

), however spring and fall values were within observation uncertainty and the timing of maximum GPP was also correct.

Pacific Northwest climate is characterized by wet springs and dry summers leading to late summer drought. Climate models predict this trend will become more

63 enhanced with warmer and wetter winters followed by drier summers (Mote and Salathé

2010). There has been concern that overestimation of modeled springtime NPP and GPP can lead to overestimation of biomass resulting in overestimation of fire burn area and severity (Rogers et al. 2011) or drought-induced mortality. Prediction of increased fire and mortality is important for land management planning and placement of forest health treatments to reduce risk. Because CLM4 is not overestimating spring GPP, we feel the seasonal dynamics are adequate for making future predictions without the risk of overestimating fire and mortality. Based on the historical evaluation, we are also confident CLM4 is adequately representing spatial carbon biomass and production patterns in order to make predictions about changes to future carbon dynamics following the implementation of varying management scenarios.

Model simulations of the management scenarios are summarized in Table 4.2 and treatment specifications, removals, and the available treatment areas are shown in Figure

4.1 and Table 4.3. We implemented varied combinations of thinning, clearcut, salvage, and no harvest management strategies. Grid cells were identified for treatments harvests based on their current productivity, mean fire return interval (MFRI), projected susceptibility to insect-related mortality, and stand density. Forest inventory data were used to calculate productivity and stand density, MFRI was obtained from the

LANDFIRE database (USGS 2009), and potential insect mortality areas were identified using published maps of current insect mortality (Raffa et al. 2008) and predicted potential vulnerability of tree species to climate change (Coops and Waring 2011). Areas with low MFRI (< 40 years) and high stand density due to fire suppression were chosen for the thinning scenarios in the East Cascades and Blue Mountains and in high density areas of the West Cascades. The grid cells selected for simulated thinning were limited to areas where average FIA derived stand densities are greater than 300 trees per hectare and capable of producing 10 Mg of merchantable wood per hectare per year. Two areas for salvage harvest following insect-related mortality were selected from the East

Cascades and the Blue Mountains. These areas represent a sub-set of the current and projected forested area affected by insect-related mortality. Finally, portions of the Coast

Range, Klamath Mountains, and West Cascades were selected for clearcut treatments.

64

Again, only grid cells capable of producing 10 Mg of merchantable wood per hectare per year were treated. Old growth reserves and roadless areas, about 20% of the forested area, were excluded from all treatments. The treatments replaced the current harvest rates for each identified grid cell to ensure there was not spatial overlap of current harvest rates with BAU conditions. For grid cells not selected, the current harvest rates (BAU) remained the same. BAU harvest rates were prescribed based on harvest since the Pacific

Northwest Forest Plan was implemented in 1990.

To implement the harvest treatments, we modified the harvest rates in the dynamic land use file used by CLM4 as the prescribed wood removal rate per year. The harvest rates in CLM4 were designed for much larger grid cell sizes and actual removal rates vary from the annual harvest rate prescribed. The result is an underestimation of harvest removals. For each grid cell treated, clearcut harvests removed 80-90% of the tree biomass for a prescribed 95% harvest rate, salvage harvests removed 65-75% for a prescribed 90% harvest rate, and thinning treatments removed 30-40% for a prescribed

50% harvest rate. While these removals are lower per unit area than what is typical for clearcuts in the region and fuels reduction treatments to reduce the risk of crown fire

(Campbell et al. 2012), the total biomass removals are reasonable compared to historical harvest removals. Total thinning, clearcut, and salvage treatment areas amounted to 36,

21, and 5% of total forested area in the region, respectively. Treatment years were staggered so that no more than 2% of the total forested area was treated each year to be consistent with average national and regional harvest rates and current mill capacity in the study region (Table 4.3). For the clearcut and salvage harvests, the plant functional types remained constant for each grid cell treated. Thinning opens up forest canopies allowing understory shrubs to grow, effectively transferring overall stand productivity to a different PFT (Campbell et al. 2009). To more adequately represent vegetation dynamics following thinning, the PFT grid cell weights were modified following a thinning harvest by transferring 20% of the forested PFTs to the shrub PFTs and transferred back to the forested PFTs in later years over a 20 year period. Finally, to simulate insect outbreak and mortality before salvage harvests we ran a single year where the mortality rate was set to 95% in the identified susceptible areas as described by

Edburg et al., (2011). For the No Harvest scenario, all harvest rates were set to zero.

Because we include the insect-killed areas in the no harvest scenario, we were also able

65 to calculate the net carbon balance if the dead trees are allowed to decompose versus salvage harvesting for bionenergy and wood products.

Model forcing datasets

Offline simulations of CLM4 are typically forced with the NCEP reanalysis dataset (Qian et al. 2006) provided by NCAR. While this dataset includes the required climate and forcing variables at a sub-daily timestep, the spatial resolution (~2 degrees or

200 km) is not adequate for sub-regional simulations in Oregon, where there are strong climatic and vegetation gradients. For this reason, we forced the model with a 1/8th degree regional dataset assembled by the Climate Impacts Group at the University of

Washington (http://www.cses.washington.edu/data/ipccar4/). Daily gridded historical

(1900-2000) and future (2000-2100) data are provided from three different models participating in the IPCC assessment reports. The regional downscaled dataset includes daily precipitation, minimum and maximum temperature, and wind speed. Because

CLM4 also requires shortwave radiation and relative humidity we calculated values incorporating algorithms from DAYMET (Thornton et al. 1997) and methods for subdaily calculations as described by (Appendix B; Göeckede et al. 2010).We chose to use the dataset that was downscaled from ECHAM A2 6 hourly simulation output as it presents the ‘middle of the road’ scenario for future projections.

The IPCC requested from the scientific communities new scenarios

(representation concentration pathways; RCPs) for modeling future climate change according to the radiative forcing expected based on socioeconomic, technological, and biophysical parameters (IPCC 2007). We chose to use the RCP4.5 pathway that leads to a radiative forcing level of 4.5 W m

-2

(approx. 550 ppm atmospheric CO

2

concentration) by the end of the century as it represents a ‘middle of the road’ scenario consistent with the climate forcing data and current harvest rates. Because atmospheric CO

2

has almost reached 400ppm already ( http://www.esrl.noaa.gov/gmd/ccgg/trends/history.html

), a higher scenario (RCP8.5) may turn out to be more realistic, but we chose the moderate

level to be consistent with remaining assumptions. Our BAU scenario does not include

66 increases to current harvest rates in the region which declined on publically owned forestland in the early 1990’s, but has remained relatively stable since 2000 (Turner et al.

2012). NCAR supplies half degree gridded aerosol, land cover change, and nitrogen deposition forcing datasets for future simulations according to the RCP4.5 scenario. We used the NCAR spatial interpolation tools to re-grid the datasets to our regional 1/8 th degree resolution and constructed a global (single value per year) CO

2 forcing dataset per the RCP4.5 trajectory where CO

2

increases to ~550 ppm.

Life cycle assessment (LCA)

In Chapter 2 and Hudiburg et al. (2011), we developed the LCA approach to quantifying net CO

2

emissions from wood use. It is detailed in Appendix A. Net CO

2 emissions to the atmosphere are determined using life cycle assessment (LCA) of wood use. Studies have identified ranges of efficiencies and energy inputs for realistic LCAs that also consider technological advances for future wood bioenergy (e.g. Mitchell et al.

2012, Winford and Gaither Jr. In Press). The values associated with the efficiency of wood product and energy conversion, energy inputs for harvest, transport, and manufacturing, and displacement of fossil fuel emissions depend on the site location, facilities available, fossil fuel source, wood products produced, land fill input rates, wood product substitution and many others (Appendix A).

Life-cycle assessment of forest carbon removals includes forestry-related sinks and sources of carbon to and from the atmosphere and the associated impact on total fossil fuel emissions (FFE). For each scenario, the net flux of carbon from or to the atmosphere (net carbon emissions; Net C e

) over 90 years (2010-2100) was calculated as the difference between the sources and the sinks following this process:

Net carbon emissions (Net C e

) = NBP + Total Harvest – WD1 – WD2 – Wood

Industry FFE – Bioenergy Emissions + Bioenergy Substitution + FF Well-To-Tank

Emissions displacement + Wood Substitution (Eq. 1)

Where, WD1 is the wood lost during manufacturing processes, WD2 is the wood decomposed over time from product use and wood substitution is included with the

assumption that there is an increased demand for wood supply. Total harvest is added

67 back to NBP to represent the theoretical amount of wood that could be stored in a wood product or converted to bioenergy if the process was 100% efficient. The WD1 variable accounts for the wood losses because wood product conversion is not 100% efficient although up to 25% of harvest and mill residues are used internally at some processing facilities for bioenergy offsetting a portion of the losses (Lippke et al. 2011). We incorporated potential mill use of current harvest residues as part of the BAU scenario.

This is different from the LCA described in Hudiburg et al. (2011) where current use of forest residues for bioenergy were not included as part of the BAU net emissions calculations. This does not reduce the WD1 term in the equation (the wood is still combusted and emissions still occur), but it increases the bioenergy substitution for fossil fuel emissions. Net C e

(net emissions from LCA equation) values are positive for carbon sinks and negative for carbon sources. In the figures and tables, ‘delta Net C e

’ refers to the difference between the management scenario Net C e

values and the BAU value. Net

C e

can be positive in both cases, but negative ‘delta Net C e

’ values indicate increased emissions (or decreased uptake) compared to BAU. In other words, the sink strength is weakened.

To quantify the change in Net C e

for each scenario, we calculate the difference between each scenario and the BAU Net C e

. The physical sinks are forest net uptake

(NBP) and wood products (Harvest) and the added virtual sinks of bioenergy and wood product substitution (FF Substitution). We exclude imports and exports from the study region since we are only interested in quantifying domestic wood production emissions and exports are less than 1% of harvested merchantable wood

(http://www.fs.fed.us/pnw/ppet/). FFE and ‘Emissions’ variables in the equation include release of carbon from woody biomass combustion and FFE associated with harvest

(Sonne 2006), transport of both harvested material and end-use products (Evans and

Finkral 2009, Heath et al. 2010), and processing and manufacturing of wood products

(Heath et al. 2010) and bioenergy (Winford and Gaither Jr. In Press). ‘Decomposition’ includes loss of material through decomposition or combustion during the manufacturing

of wood products (Smith et al. 2006), and the percentage of wood products that are expected to no longer be in-use at the end of the treatment period (Smith et al. 2006).

68

Biomass utilized for wood products can end up in a long term storage product

(structural wood) or a short term product (paper). Some wood product carbon reenters the atmosphere through rapid (paper) or slow (wood) decomposition or combustion while some is eventually disposed in landfills where it is very slowly decomposed. West Coast harvests generate merchantable bole wood at rates of 50-60% of the total wood harvested

(Harmon et al. 1996) and decay at a net rate of 1% per year (Harmon and Marks 2002,

Smith et al. 2006) after accounting for the portion stored in landfills. Using values provided by (Smith et al. 2006), we determined the amounts of long and short term wood products that could be generated by the merchantable wood harvested accounting for the losses along the way using the net decay rate. The remaining non-merchantable wood from harvest was used for combined heat and power (CHP) bioenergy. We also accounted for the associated emissions for both conversion to wood chips and the combustion emissions.

Fossil fuel substitution with bioenergy was calculated as biomass combustion for

CHP compared to fossil fuel sources. Woody biomass provides less energy per unit of carbon emitted than fossil fuels (i.e. wood has an energy content of 20 GJ per ton versus

35.5 GJ per ton in coal and 58 GJ per ton in natural gas) because fossil fuels have a lower heating value (Wright et al. 2006). The conversion efficiency of biomass to CHP compared to the reference fossil fuel source ranges from 20-80% depending on the power plant and the fossil fuel source being replaced (Mitchell et al. 2012). The US average conversion efficiency is 51% given a combination of low to highly efficient plants and the US mix of fossil fuel CHP production (coal, natural gas and petroleum/oil). State annual fossil fuel emissions, energy sources, and consumption were acquired from the

Oregon Department of Energy

(http://www.oregon.gov/energy/pages/oregons_electric_power_mix.aspx). The Oregon average conversion efficiency given the state energy mix is very close to the US average at 50%. This was also an improvement over the LCA used in Hudiburg et al. (2011) where the fossil fuel source replaced was petroleum/oil only.

There are also emissions associated with crude extraction and manufacturing, sometimes called the wells-to-tank emissions (WTT). Fossil fuel LCA total emissions

69

(wells to wheels; WTW) include both WTT and tank-to-wheels (TTW) emissions. The amount of carbon emitted per unit of fossil fuel energy varies widely by source fuel, but average WTT emissions are approximately 15% (ICCT 2010) of total emissions (WTW), or 12 g CO

2

per MJ of energy. We have included these emissions in the Wood Industry

FFE and we have added a WTT displacement benefit along with the bioenergy substitution benefit.

Finally, we add potential wood product substitution benefits for replacement of fossil fuel derived products. Wood product substitution for a 50/50 mix of aluminum and steel used in residential American housing generates a 36% reduction in fossil fuel emissions (Upton et al. 2008) and 26% for concrete (Lippke et al. 2010). We assumed these rates will continue into the future for new residential housing and applied a 36% wood substitution benefit of the final structural wood product pool to represent optimal substitution rates.

Results and Discussion

What are the interactive effects of changing climate, increasing atmospheric CO

2

, and land use change on net ecosystem production (NEP)?

At the end of the 21 st

century, predicted regional mean annual air temperature is

11.5 °C, an increase of 3.5 °C over the mean annual temperature in 2010 (Figure 4.2a).

Annual precipitation averages 1050 mm in the last five years of the century, increasing only 25 mm over early century averages. However, the predicted precipitation patterns vary by ecoregion with a decline in precipitation of about 40 mm yr

-1

in the mesic Coast

Range, West Cascades, and Klamath Mountains versus a 115 mm yr

-1

increase in the drier East Cascades and Blue Mountains. Interannual variability in precipitation is high for the entire simulation period. Seasonal variation of precipitation on monthly GPP revealed slightly increased GPP in summers following wet winters in the semi-arid Blue

Mountains and East Cascades, but no change in the mesic ecoregions (Appendix B Figure

B6).

70

Regional relative humidity remains constant overtime (Appendix B, Figure B5).

Predicted atmospheric CO

2

concentrations are 550 ppm and nitrogen deposition varies from 2.2 kg N ha

-1

yr

-1

in the Coast Range, West Cascades, and Klamath Mountains to

3.3 kg N ha

-1

yr

-1

in the East Cascades and Blue Mountains. This is a change of < 1 kg N ha

-1

yr

-1

for the study region and it is less than typical values for temperate forests receiving high levels of nitrogen deposition from industrial and agricultural sources (>

5.5 kg N ha

-1

yr

-1

; Janssens et al. 2010).

Predicted regional net ecosystem production (NEP) in 2100 is 13.7 Tg C yr

-1

(107 g C m

-2

yr

-1

) for BAU conditions compared with the current NEP of 13.2 ± 1.6 (103 g C m

-2 yr

-1

) indicating that the effects of changing climate, nitrogen deposition, and increasing CO

2

concentrations offset each other after the 90 year period (bright green line scenario in Figure 4.2 and Table 4.4). Nitrogen deposition, precipitation patterns, and relative humidity do not change significantly and therefore changes in predicted carbon fluxes are due to climate warming ( (Figure 4.2a), CO

2

fertilization, and/or disturbance.

Increases in BAU NPP (Figure 4.2b) are accompanied by increases to heterotrophic respiration (R h

), resulting in no change in NEP (Figure 4.2c).

The modeled response to warming and rising CO

2

is not surprising and is in line with previous work (Thornton et al. 2009). Net carbon uptake has been shown to increase in response to elevated CO

2

concentrations in laboratory-grown tree seedlings (Curtis and

Wang 1998) and experiments on whole forest canopies (DeLucia et al. 2005). Warming has been shown to increase decomposition which increases nitrogen mineralization, followed by increases in NPP. However, eventually nitrogen availability limits carbon assimilation (Thornton et al. 2007). Predicted soil mineral nitrogen is maintained for the first 40-50 years and then declines (Figure 4.2d). Finally, a simulation where only climate was allowed to vary (black lines in Figure 4.2), shows increases in NPP followed by decline, but R h

only increases over time

.

In Pacific Northwest forests, our results support the hypothesis that increases in NPP and R h

due to climate warming are enhanced by CO

2 fertilization and warming until nitrogen limitation occurs and carbon uptake declines following a peak in 2030 and then drops below its current value (Fig 4.2c, red line).

71

As expected, the thinning and clearcut strategies that increase harvest compared to

BAU, result in consistently higher regional NEP (Figure 4.4) although they follow the same general pattern as BAU over time (Figure 4.3). This is primarily due to reductions in heterotrophic respiration (R h

) because of a declining woody detritus pool creating an imbalance between NPP and R h

(Figure 4.3). The predicted decline in R h

compared to

BAU is due to decreases in decomposable material (i.e. woody detritus; Table 4.4) because the bioenergy harvests remove harvest residues that would have otherwise decomposed onsite and reduce inputs to the woody detritus pool by removing a portion of the standing biomass. However, the wood removed is either combusted, stored in a wood product, or decomposed offsite adding to carbon emissions in the life cycle assessment.

NBP averages 5.5 Tg C yr

-1

(38 ± 8 g C m

-2

yr

-1

) under a BAU management scenario that maintains present harvest rates over the next 90 years. NBP declines from an average of 6.2 Tg C yr

-1

in the first half of the century to 4.5 Tg C yr

-1

in the last part of the century. NBP changes with increases or decreases to NEP, fire, or harvest. For the

BAU scenario, we find no significant change to NEP over time and harvest rates are purposely held constant. However, projected burn area doubles by the end of the century increasing fire emissions from an average of 2.5 to 3.5 Tg C yr

-1

(Appendix B, Figure

B4) compensating for most of the decline in NBP. Historical comparisons of CLM4 simulated burn area (Appendix B, Figure B3) with other datasets (MTBS; Eidenshink, J. et al. 2007 and GFED; van der Werf, G. et al. 2010) show both over- and underestimation by CLM4. The Global Fire Emissions database (GFED) underestimated burn area by

15% for the Oregon Biscuit fire in 2002 and was consistently the lowest burn area estimate compared other models in a synthesis done by French et al. (2011). However, the predicted increase in burn area over the 21 st

century is consistent with other studies

(Rogers et al. 2011, Westerling et al. 2006).

The majority of the management scenarios cause NBP to decline by 2- 4 Tg C yr

-1 because harvest is increased removing more wood compared to BAU. For the treatments intended to prevent fire-related emissions, the reduction in predicted fire emissions is not enough to compensate for the harvest losses (Figure 4.4). The overall increases in treatment NEP compared to BAU (Table 4.4) are not sufficiently large enough to balance

the wood removals from harvest, although some of these losses are recouped offsite in

72 wood products and substitution of fossil fuels with bioenergy. The state level results indicate that the effects of changing climate, nitrogen deposition, and increasing CO

2 concentrations offset each other after the 90 year period, resulting in nearly equal NEP at the beginning and end of the 21 st

century for BAU conditions. For BAU management where harvest rates are held constant, statewide NBP still declines over the 90 year period compared to the first part of the century. This is primarily due to increased fire emissions from an increase in predicted burn area by CLM4 consistent with other studies.

The only scenario that does not reduce NBP is the No Harvest scenario where NBP increased by approximately 4 Tg C yr

-1

(30 g C m

-2

yr

-1

) despite increases in fire because harvest removals are stopped.

The thinning and clearcut bioenergy scenarios result in increased NEP compared to BAU at the end of the 21 st

century because of a decline in heterotrophic respiration

(R h

). The decline in R h

is most likely due to a decreasing woody detritus pool because bioenergy management removes the harvest residues that would otherwise decompose onsite. However, NBP is still reduced for the bioenergy management scenarios compared to BAU emphasizing that the increase in NEP and treatment reductions in predicted fire emissions do not compensate for the wood removals. The opposite is true for the No

Harvest scenario where NBP and total forest carbon are increased despite a decline in

NEP and increase in fire emissions.

How do the varying management scenarios affect carbon stocks and fluxes at different spatial scales?

For all ecoregions, NPP generally decreased for several years following the bioenergy management scenarios compared to BAU. NPP recovered between 5 – 30 years following each treatment depending on the intensity of the harvest (thinning vs. clearcut) and the ecoregion where the treatment was performed. Recovery times were generally longer for both the first thinning and the first clearcut treatments because more biomass was available for removal. Subsequent removals were less because aboveground stem biomass had not fully recovered to BAU levels (Figure 4.4e-4.8e), resulting in less

biomass available for harvest (Figure 4.4f-4.8f). This is consistent with research in

Pacific Northwest showing a decline in baseline carbon stocks over time with either increased harvest removals or decreased rotation time (Harmon and Marks 2002).

A study of a thinning chronosequence in a mixed conifer forest of the Sierra

Nevada region of California found stand level NPP recovery rates of about 80% for shrubs and trees combined and almost complete recovery of NEP after 16 years, with shrubs compensating for lost NPP due to thinning (Campbell et al. 2009). Modeled shrub

NPP increased following the thinning treatments as expected because the transfer of tree

73

NPP to shrub NPP was prescribed in the simulations by transferring a percentage of the grid cell from the tree PFT to the shrub PFT in the dynamic PFT input file. Shrub NPP declined as the percentage was slowly transferred back to the tree PFT in accordance with the observed data.

Simulated recovery after clearcut management was followed by increases in NPP compared to BAU, as expected, from a combination of the positive effect of climate and

CO

2

fertilization and young stand development following harvest in the more productive ecoregions (Coast Range and Klamath Mountains; Figure 4.5a and 4.7a) and salvage logging management in the insect-kill areas (Figure 4.4a and 4.6a). Thinning treatments are known to increase the biomass production of remaining individuals, but have little to no effect on stand level NPP because ultimately site resources limit total production (Law et al. 2012), i.e. the Law of Constant Final Yield. We found this to be true in the East

Cascade thinning treatments where NPP following the first thinning harvest returned to

BAU levels after 30 years and did not appreciably surpass BAU. The increases beyond

BAU NPP in the latter thinning treatments are associated with the regional climate and

CO

2 fertilization trends. This was not the case in the Blue Mountains where NPP in the thinning treatment exceeds BAU NPP following each disturbance. Although, there was no change in regional mean annual precipitation, precipitation in the Blue Mountains increased by 130 mm by the end of the study period suggesting that drought stress impacts on NPP were alleviated by the additional precipitation. There is some evidence from other studies that thinning relief of drought stress could result in higher stand level

NPP (McDowell et al. 2003, Kolb et al. 2007).

74

There is an increasing trend in NEP and NBP over time compared to BAU for the thinning and clearcut bioenergy management scenarios in the Blue Mountains and East

Cascades. Reasons for this again include the positive effects of CO

2

and warming, but also because of a decline in R h

due to declines in detritus pools because of removals for bioenergy use. Since NEP is the balance between NPP and R h

, as respiration declines without concurrent declines in NPP, NEP increases in all management scenarios compared to BAU. Another reason for increases in NBP overtime is the decline in subsequent harvest removals (Figures 4f-8f) following the initial harvest. Prescribed harvest rates were the same for each harvest (~85% of stem biomass for clearcuts and

~35% of stem biomass for thinning), however because stem biomass carbon had not recovered to initial conditions before the second treatment, harvest removals declined.

This is important because the rotation lengths will be determined by both fire prevention measures and wood supply demands for products and energy. The rotation lengths suggested here are sustainable over time and under this climate projection and CO

2 scenario, as indicated by the recovery of NBP. However, the initial carbon debt will take longer to recoup because the carbon lost in the first treatment is not fully re-sequestered in forest biomass over the next century.

In summary, the ecoregion trends in R h

and NEP are similar to the statewide results for the management scenarios, R h

declines and NEP increases. NPP is increased above BAU in the Blue Mountains for the thinning scenario, as did precipitation, suggesting Blue Mountain NPP was limited by drought stress. NBP declines compared to

BAU over the 90 year period for each of the bioenergy scenarios, but the recovery rates vary by harvest intensity. There is some evidence that the combination of increased NEP due to climate warming and CO

2

fertilization and increased NEP due to reduced R h

in the scenarios is contributing to an increase in NBP compared to BAU at the end of the 21 st century.

Does bioenergy management increase or decrease net CO

2

emissions to the atmosphere compared to BAU by the end of the 21st century? Where do net CO

2

emissions decrease compared to BAU for bioenergy management?

75

After complete life cycle assessment of the removed wood, we find that none of the bioenergy management scenarios reduce emissions to the atmosphere by the end of the 21 st

century. Overall, the more intensive thinning and clearcut scenarios increase state-wide emissions to the atmosphere compared to BAU over the 90 year period whereas the less intensive salvage logging scenario results in no change.

The bioenergy management scenarios increase landscape level net CO

2

emissions

(Net C e in equation 1 is decreased) to the atmosphere over the next century compared to

BAU, especially in the Coast Range and West Cascades (Figure 4.9) compared to BAU.

The No Harvest scenario is the only scenario that significantly decreased overall emissions by 2100 except for in the Klamath Mountains. For the No Harvest scenario,

Net C e

is equivalent to NBP because there are no associated wood product or bioenergy emissions or substitutions. In the Coast Range and West Cascades ecoregions, the decline in harvest rates between BAU and No Harvest result in large increases in NBP compared to BAU and the offsite wood product sinks and bioenergy substitutions are not large enough to compensate for the difference. However this is not the case in the Klamath

Mountains where a relatively small BAU harvest exits; the small harvest rate combined with offsite wood product and bioenergy substitutions amounts to lower carbon emissions for BAU compared to a No Harvest scenario with no offsite wood product sink or bioenergy substitutions. These results suggest that for the Klamath Mountains, there exist specific combinations of harvest, wood products, and bioenergy management that reduce emissions compared to management plans with no harvest at all. However, at the statelevel, a No Harvest scenario results in overall emissions reductions despite the increase in the Klamath Mountains.

Given the simulated extent and low intensity, salvage harvesting has almost no significant effect on emissions in the treated ecoregions and at the state level, indicating this management strategy is an acceptable bioenergy management strategy from an emissions perspective as long as this low intensity of salvage is applied. Simulated salvage harvests are small, not repeated, and the combined increase in NEP (reduced R h

) and offsite sinks and substitutions do compensate for the harvest losses in this experiment overtime. In reality, for those regions where insect- and disease-related mortality are

76 occurring or predicted to increase, the salvage logging treatment would probably have to be repeated after forest regrowth and subsequent mortality. This would change the salvage logging treatment impact and increase the amount of time required to reduce emissions. However, these results do support the hypothesis that bioenergy management should target those areas already experiencing large decrease in net carbon uptake.For the west coast region, net uptake would have to decline by 40% for short-term emissions reductions, but by as little as 10-15% in the Blue Mountains and East Cascades

(Hudiburg et al. 2011; Ch. 2).

In the Blue Mountains, the thinning scenario decreases net emissions (positive values for Net C e

) compared to BAU at the end of the 21 st

century, indicating the longterm effects of bioenergy management combined with environmental change will decrease CO

2

emissions to the atmosphere. The Blue Mountains are a semi-arid ecoregion where productivity is much lower than the mesic ecoregions. However it takes nearly 60 years for the Net C e

trend to become mostly positive, primarily because of the subsequent thinning removals. This trend is not as evident in the East Cascades (Figure

4.13a) and West Cascades (Figure 4.15a) where the change in Net C e

does not become positive until after the second thinning rotation (Figure 4.13b and Figure 4.15b). These results suggest that it will take longer in both the East and West Cascades to realize overall net reductions in CO

2

emissions, however the amount of time cannot be determined because is it longer than the simulation period.

Net carbon emissions in the clearcut scenario are increased in the Coast Range and Klamath Mountains (Figure 4.12a and 4.14a) for about 35 years after the first clearcut, with less impact after the second harvest because the second harvest is smaller.

Net C e in the No Harvest scenario is greater than BAU for the majority of the simulation period for the three mesic ecoregions resulting in less cumulative emissions compared to BAU after 90 years, except in the Klamath Mountains. The reason for the increase in the Coast Range and West Cascades is because of reduced harvest compared to BAU. However, Net C e

declines (Figure 4.12b, 4.14b, and 4.15b) and is no longer reducing CO

2

emissions compared to BAU after 50 – 60 years in the three mesic ecoregions. NEP declines over time for these ecoregions in the No Harvest scenario

77 because of increases in R h and eventually the reductions in harvest do not compensate for the reductions in NEP. However, there is evidence that NEP can remain positive

(Luyssaert et al. 2008) in old-growth temperate forests and with no harvest or large increases in fire, Net C u

would also remain positive in the absence of catastrophic disturbance.

Overall, the more intensive thinning and clearcut scenarios increase state-wide emissions to the atmosphere compared to BAU over the 90 year period whereas the less intensive salvage logging scenario does not. In the semi-arid Blue Mountains ecoregion, net uptake is greater than BAU at the end of the 21 st

century indicating the long-term effects of bioenergy management combined with environmental change will decrease

CO

2

emissions to the atmosphere. In all ecoregions except for the Klamath Mountains, the No Harvest scenario reduced CO

2 emissions compared to BAU over the 90 year period. In the Klamath Mountains, small BAU harvest rates combined with offsite wood product and bioenergy substitutions amounted to lower carbon emissions for BAU compared to a No Harvest scenario with no offsite wood product sink or bioenergy substitutions. These results suggest that for the Klamath Mountains, there exists a specific combination of harvest, wood products, and bioenergy management that reduce emissions compared to management plans with no harvest at all.

Conclusions

At the end of the 21 st

century, predicted regional net ecosystem production (BAU

NEP) does not change compared with the current NEP indicating compensating effects of changing climate, nitrogen deposition, and increasing CO

2

concentrations over the next

90 years. In other words, statewide rates of net carbon uptake under business-as-usual conditions are not reduced compared to current conditions. Despite no change to NEP or harvest rates, BAU NBP still declines by ~2 Tg C yr

-1

by the end of the 21 st

century, primarily due to increases in fire emissions and fire related mortality predicted by CLM4 under future climate conditions.

Implementation of bioenergy management strategies that increase harvest compared to BAU (thinning and clearcut scenarios), do not result in net reductions to

carbon emissions over the next 90 years despite inclusion of all wood product and bioenergy substitutions. The management associated increases in net carbon uptake

(NEP), reduction in fire emissions, and associated wood product sinks and substitutions are not sufficiently large enough to balance the life cycle emissions from product acquisition, conversion, decomposition, and combustion for energy by the end of the 21 st

78 century.

The dominant trend at the regional level varies at different spatial resolutions and in some ecoregions (Blue Mountains and East Cascades), net carbon uptake is beginning to improve over BAU at the end of the 90 year period suggesting bioenergy management could lead to net reductions over a longer time period (> 100 years). If the bioenergy treatments were only implemented in these two ecoregions total state-wide emissions would still increase by 10 Tg C by the end of the 21 st

century compared to 140 Tg C if all ecoregions are treated. Salvage logging has almost no significant effect on emissions in the treated ecoregions and at the state level, however the salvage logging treatments are small compared to the other treatments (<4% increase in removals compared to BAU harvest). Finally, in the Klamath Mountains, there is evidence that the current level of harvest, wood products, and bioenergy management reduces emissions compared to management plans with no harvest at all.

Longer simulations and better fire and insect related mortality prediction will be required to identify areas where long term bioenergy management strategies will be most effective. Also, the simulations do not include a mechanism for PFTs to change overtime

(i.e. dynamic vegetation model) due to stress caused by disturbance factors such as insects, disease and fire or drought. Also, the annual scenario harvest rates presented here are not necessarily reflective of industry operations, which would require a steady supply of wood over space and time. Current mill capacity in Oregon is capable of processing the suggested harvest increases in merchantable wood, but there are only a few wood biomass CHP facilities in operation in Oregon although construction is planned for more

(Siemers 2011). Finally, inclusion of carbon removals in runoff and flood events (DOC; dissolved organic carbon) also needs to be incorporated in the CLM4 carbon budget as they can increase by 50-200% in the years following harvest (Morris 2009).

79

Policy and management plans need to consider the land-based sink in addition to off-site wood usage when evaluating options for bioenergy from forest biomass. In this study, none of the scenarios reduce CO

2

emissions and there is only one treatment where bioenergy management results in no change in emissions. This is another example of

“slow in and fast out” where consumption (increased emissions from bioenergy) exceeds growth (Schultz et al. 2012). In other words, forest carbon that took decades to centuries to accumulate is rapidly lost to the atmosphere through harvest for bioenergy.

References

Bonan, G. B., P. J. Lawrence, K. W. Oleson, S. Levis, M. Jung, M. Reichstein, D. M.

Lawrence, and S. C. Swenson. 2011. Improving canopy processes in the

Community Land Model version 4 (CLM4) using global flux fields empirically inferred from FLUXNET data. J. Geophys. Res. 116:G02014.

Breshears, D. D., O. B. Myers, C. W. Meyer, F. J. Barnes, C. B. Zou, C. D. Allen, N. G.

McDowell, and W. T. Pockman. 2008. Tree die-off in response to global changetype drought: mortality insights from a decade of plant water potential measurements. Frontiers in Ecology and the Environment 7:185-189.

Campbell, J., G. Alberti, J. Martin, and B. E. Law. 2009. Carbon dynamics of a ponderosa pine plantation following a thinning treatment in the northern Sierra

Nevada. Forest Ecology and Management 257:453-463.

Campbell, J. L., M. E. Harmon, and S. R. Mitchell. 2012. Can fuel-reduction treatments really increase forest carbon storage in the western US by reducing future fire emissions? Frontiers in Ecology and the Environment 10:83-90.

Coops, N. C., and R. H. Waring. 2011. Estimating the vulnerability of fifteen tree species under changing climate in Northwest North America. Ecological Modelling

222:2119-2129.

Curtis, P. S., and X. Z. Wang. 1998. A meta-analysis of elevated CO2 effects on woody plant mass, form and physiology. Oecologia 113:299-313.

DeLucia, E. H., D. J. Moore, and R. J. Norby. 2005. Contrasting responses of forest ecosystems to rising atmospheric CO2: Implications for the global C cycle.

Global Biogeochemical Cycles 19:doi:10.1029/2004GB002346.

80

Eidenshink, J. et al. A project for monitoring trends in burn severity. Fire ecology 3, 3-20

(2007).

Evangelista, P. H., S. Kumar, T. J. Stohlgren, and N. E. Young. 2011. Assessing forest vulnerability and the potential distribution of pine beetles under current and future climate scenarios in the Interior West of the US. Forest Ecology and Management

262:307-316.

Evans, A., and A. Finkral. 2009. From renewable energy to fire risk reduction: a synthesis of biomass harvesting and utilization case studies in US forests. GCB

Bioenergy 1:211-219.

French, N. H. F., W. J. de Groot, L. K. Jenkins, B. M. Rogers, E. Alvarado, B. Amiro, B. de Jong, S. Goetz, E. Hoy, E. Hyer, R. Keane, B. E. Law, D. McKenzie, S. G.

McNulty, R. Ottmar, D. R. Pérez-Salicrup, J. Randerson, K. M. Robertson, and

M. Turetsky. 2011. Model comparisons for estimating carbon emissions from

North American wildland fire. J. Geophys. Res. 116:G00K05.

Harmon, M. E., J. M. Harmon, W. K. Ferrell, and D. Brooks. 1996. Modeling carbon stores in Oregon and Washington forest products: 1900–1992. Climatic Change

33:521-550.

Harmon, M. E., and B. Marks. 2002. Effects of silvicultural practices on carbon stores in

Douglas-fir – western hemlock forests in the Pacific Northwest, U.S.A.: results from a simulation model. Canadian Journal of Forest Research/Revue Canadienne de Recherche Forestiere 32:863-877.

Heath, L. S., V. Maltby, R. Miner, K. E. Skog, J. E. Smith, J. Unwin, and B. Upton.

2010. Greenhouse Gas and Carbon Profile of the U.S. Forest Products Industry

Value Chain. Environmental Science & Technology 44:3999-4005.

Hudiburg, T., B. E. Law, D. P. Turner, J. Campbell, D. Donato, and M. Duane. 2009.

Carbon dynamics of Oregon and Northern California forests and potential landbased carbon storage. Ecological Applications 19:163-180.

Hudiburg, T. W., B. E. Law, C. Wirth, and S. Luyssaert. 2011. Regional carbon dioxide implications of forest bioenergy production. Nature Clim. Change 1:419-423.

Hurtt, G. C., S. Frolking, M. G. Fearon, B. Moore, E. Shevliakova, S. Malyshev, S. W.

Pacala, and R. A. Houghton. 2006. The underpinnings of land-use history: three centuries of global gridded land-use transitions, wood-harvest activity, and resulting secondary lands. Global Change Biology 12:1208-1229.

ICCT. 2010. Carbon intensity of crude oil in Europe. International Council on Clean

Transportation, Washington D.C.

IPCC. 2007. IPCC expert meeting report: Technical Summary, Towards new scenarios for analysis of emissions, climate change, impacts, and response strategies.

Intergovernmental Panel on Climate Change, Cambridge.

81

Janssens, I. A., W. Dieleman, S. Luyssaert, J. A. Subke, M. Reichstein, R. Ceulemans, P.

Ciais, A. J. Dolman, J. Grace, G. Matteucci, D. Papale, S. L. Piao, E. D. Schulze,

J. Tang, and B. E. Law. 2010. Reduction of forest soil respiration in response to nitrogen deposition. Nature Geosci 3:315-322.

Jenkinson, D. S., K. Goulding, and D. S. Powlson. 1999. Nitrogen deposition and sequestration. Nature 400:629.

Kolb, T. E., J. K. Agee, P. Z. Fulé, N. G. McDowell, K. Pearson, A. Sala, and R. H.

Waring. 2007. Perpetuating old ponderosa pine. Forest Ecology and Management

249:141-157.

Kurz, W. A., G. Stinson, G. J. Rampley, C. C. Dymond, and E. T. Neilson. 2008. Risk of natural disturbances makes future contribution of Canada's forests to the global carbon cycle highly uncertain. Proceedings of the National Academy of Sciences

105:1551-1555.

Law, B., T. Hudiburg, and S. Luyssaert. 2012. Thinning Effects on Forest Productivity:

Consequences of Preserving Old Forests and Mitigating Impacts of Fire and Drought.

Plant Ecology & Diversity doi: 10.1080/17550874.2012.679013

Lawrence, D., K. W. Oleson, M. G. Flanner, P. E. Thorton, S. C. Swenson, P. J.

Lawrence, X. Zeng, Z.-L. Yang, S. Levis, K. Skaguchi, G. B. Bonan, and A. G.

Slater. 2011. Parameterization Improvements and Functional and Structural

Advances in Version 4 of the Community Land Model. J. Adv. Model. Earth Syst.

3:27.

Lippke, B., E. Oneil, R. Harrison, K. Skog, L. Gustavsson, and R. Sathre. 2011. Life cycle impacts of forest management and wood utilization on carbon mitigation: knowns and unknowns. Carbon Management 2:303-333.

Lippke, B., J. Wilson, J. Meil, and A. Taylor. 2010. Characterizing the importance of carbon stored in wood products. Wood and Fiber Science 42:5-14.

Luyssaert, S., E. D. Schulze, A. Borner, A. Knohl, D. Hessenmoller, B. E. Law, P. Ciais, and J. Grace. 2008. Old-growth forests as global carbon sinks. Nature 455:213-

215.

82

Malmsheimer, R. W., J. L. Bowyer, J. S. Fried, E. Gee, R. L. Izlar, R. A. Miner, I. A.

Munn, E. Oneil, and W. C. Stewart. 2011. Managing Forests because Carbon

Matters: Integrating Energy, Products, and Land Management Policy. Journal of

Forestry 109:S7-S51.

McDowell, N., J. R. BROOKS, S. A. FITZGERALD, and B. J. BOND. 2003. Carbon isotope discrimination and growth response of old Pinus ponderosa trees to stand density reductions. Plant, Cell and Environment 26:631-644.

McDowell, N. G., C. D. Allen, and L. Marshall. 2010. Growth, carbon-isotope discrimination, and drought-associated mortality across a Pinus ponderosa elevational transect. Global Change Biology 16:399-415.

McKechnie, J., S. Colombo, J. Chen, W. Mabee, and H. L. MacLean. 2011. Forest

Bioenergy or Forest Carbon? Assessing Trade-Offs in Greenhouse Gas Mitigation with Wood-Based Fuels. Environmental Science & Technology 45:789-795.

Mitchell, S. R., M. E. Harmon, and K. E. B. O'Connell. 2012. Carbon debt and carbon sequestration parity in forest bioenergy production. GCB Bioenergy:n/a-n/a.

Morris, D. M. 2009. Changes in DOC and DON fluxes in response to harvest intensity of black-spruce-dominated forest ecosystems in northwestern Ontario. Canadian

Journal of Soil Science 89:67-79.

Morris, J. 2010. Bury or Burn North America MSW? LCAs Provide Answers for Climate

Impacts & Carbon Neutral Power Potential. Environmental Science &

Technology 44:7944-7949.

Mote, P. W., and E. Salathé. 2010. Scenarios: Future Climate in the Pacific Northwest.

University of Washington, Seattle.

Raffa, K. F., B. H. Aukema, B. J. Bentz, A. L. Carroll, J. A. Hicke, M. G. Turner, and W.

H. Romme. 2008. Cross-scale Drivers of Natural Disturbances Prone to

Anthropogenic Amplification: The Dynamics of Bark Beetle Eruptions.

Bioscience 58:501-517.

Rogers, B., R. Neilson, R. Drapek, J. Lenihan, J. Wells, D. Bachelet, and B. Law. 2011.

Impacts of climate change on fire regimes and carbon stocks of the U.S. Pacific

Northwest. J. Geophys. Res. 116 G03037.

Routa, J., S. KellomÄKi, A. KilpelÄInen, H. Peltola, and H. Strandman. 2011. Effects of forest management on the carbon dioxide emissions of wood energy in integrated production of timber and energy biomass. GCB Bioenergy 3:483-497.

83

Schulze, E. D., W. D. Vries, M. Hauhs, K. Rosén, L. Rasmussen, O.-C. Tann, and J.

Nilsson. 1989. Critical loads for nitrogen deposition in forest ecosystems. Water,

Air, and Soil Pollution 48:451-456.

Schulze, E.-D., C. Körner, B. E. Law, H. Haberl, and S. Luyssaert. 2012. Large-scale bioenergy from additional harvest of forest biomass is neither sustainable nor greenhouse gas neutral. GCB Bioenergy:n/a-n/a.

Siemers, E. 2011. Northwest Energy plans $300M Oregon biomass investment. Pages http://www.sustainablebusinessoregon.com/articles/2011/2003/northwest-energyplans-2300m-oregon.html

in Portland Business Journal, Portland.

Smith, J. E., L. Heath, K. E. Skog, and R. Birdsey. 2006. Methods for Calculating Forest

Ecosystem and Harvested Carbon with Standard Estimates for Forest Types of the

United States Gen. Tech. Rep. NE-343. U.S. Department of Agriculture, Forest

Service, Northeastern Research Station, Newtown Square, PA.

Sonne, E. 2006. Greenhouse Gas Emissions from Forestry Operations. J. Environ. Qual.

35:1439-1450.

Thornton, P. E., J.-F. o. Lamarque, N. A. Rosenbloom, and N. M. Mahowald. 2007.

Influence of carbon-nitrogen cycle coupling on land model response to CO2 fertilization and climate variability. Global Biogeochem. Cycles 21.

Thornton, P. E., and N. E. Zimmermann. 2007. An improved canopy integration scheme for a land surface model with prognostic canopy structure. Journal of Cllimate

20:3902-3923.

Thornton, P. E., S. C. Doney, K. Lindsay, J. K. Moore, N. Mahowald, J. T. Randerson, I.

Fung, J. F. Lamarque, J. J. Feddema, and Y. H. Lee. 2009. Carbon-nitrogen interactions regulate climate-carbon cycle feedbacks: results from an atmosphereocean general circulation model. Biogeosciences 6:2099-2120.

Turner, D., W. D. Ritts, Z. Yang, R. E. Kennedy, W. B. Cohen, M. V. Duane, P. E.

Thornton, and B. E. Law. 2012. Decadal trends in net ecosystem production and net ecosystem carbon balance for a regional socioecological system Forest Ecology and Management.

Upton, B., R. Miner, M. Spinney, and L. S. Heath. 2008. The greenhouse gas and energy impacts of using wood instead of alternatives in residential construction in the

United States. Biomass and Bioenergy 32:1-10.

84

USGS. 2009. The National Map LANDFIRE: LANDFIRE National Existing Vegetation

Type layer. Pages http://gisdata.usgs.net/website/landfire/ in U.S. Department of

Interior, Geological Survey. U.S. Department of Interior, Geological Survey. van der Werf, G. R., J. T. Randerson, L. Giglio, G. J. Collatz, M. Mu, P. S. Kasibhatla, D.

C. Morton, R. S. DeFries, Y. Jin, and T. T. van Leeuwen. 2010. Global fire emissions and the contribution of deforestation, savanna, forest, agricultural, and peat fires (1997–2009). Atmos. Chem. Phys. 10:11707-11735. van Mantgem, P. J., N. L. Stephenson, J. C. Byrne, L. D. Daniels, J. F. Franklin, P. Z.

Fule, M. E. Harmon, A. J. Larson, J. M. Smith, A. H. Taylor, and T. T. Veblen.

2009. Widespread Increase of Tree Mortality Rates in the Western United States.

Science 323:521-524.

Walker T, Cardellichio P, C. A, and E. al. 2010. Biomass sustainability and carbon policy study. NCI-2010-03 (2010). Manomet Center for Conservation Sciences, ,

Massachusetts.

Whitaker, J., K. E. Ludley, R. Rowe, G. Taylor, and D. C. Howard. 2010. Sources of variability in greenhouse gas and energy balances for biofuel production: a systematic review. GCB Bioenergy 2:99-112.

Winford, E., and J. Gaither Jr. In Press. Carbon outcomes from fuels treatment and bioenergy production in a Sierra Nevada forest Forest Ecology and Management.

Wright, L., B. Boundy, R. Perlack, S. C. Davis, and B. Saulsbury. 2006. Biomass energy data book: edition 1, 1 edition. Oak Ridge National Laboratory, Oak Ridge,

Tennessee, USA.

Zhang, F., J. M. Chen, Y. Pan, R. A. Birdsey, S. Shen, W. Ju, and L. He. 2012.

Attributing carbon changes in conterminous U.S. forests to disturbance and nondisturbance factors from 1901 to 2010. J. Geophys. Res. 117:G02021.

Table 4.1. Ecoregion characteristics including forested area, mean stand ages by ownership, dominant forest types, mean annual precipitation (MAP), mean annual temperature (MAT), and proposed bioenergy management scenarios.

Ecoregions are listed from high to low MAP.

Ecoregion

Coast Range (CR)

Forest

Hectares

Stand Age

Private/public

(% total)

2043332 (17) 34 / 75

Dominant Forest

Types

MAP

(mm yr

-1

)

1742

MAT

(C°)

11.0

Proposed

Bioenergy Management

West Cascades (WC)

Klamath Mountains

(KM)

Willamette Valley

(WV)

East Cascades (EC)

Blue Mountains (BM)

Columbia Plateau

(CP)

2693263 (22) 50 / 140

1302111

(11)

59 / 106

501793 (4) 43 / 61

2099866 (17) 64 / 94

3364151 (27) 71 /100

88922 (<1) 80 / 47

Douglas-fir, Sitka

Spruce, Redwood,

Western Red Cedar,

Fir-hemlock

Douglas-fir, Hemlock,

Mixed Conifer, Red

Fir, Western Red

Cedar

Mixed Conifer, Mixed

Evergreen, Red Fir,

Douglas-fir, Riparian,

Oak

Douglas-fir, Hemlock,

Riparian

Ponderosa Pine, Mixed

Conifer, Juniper, Pine,

Red Fir

Mixed Conifer,

Ponderosa Pine,

Juniper, Spruce-Fir

1688

1549

1280

630

552

330

8.8

11.5

11.0

9.1

7.3

9.7

45 year clearcut rotations on non-reserved secondary forestland

45 year clearcut rotations on non-reserved secondary forestland ; 30 year thinning rotations in high density stands

45 year clearcut rotations on non-reserved secondary forestland

None

30 year thinning rotations on high fire return interval forestland; Salvage harvest of insect-killed forests

30 year thinning rotations on high fire return interval forestland; Salvage harvest of insect-killed forests

None

Northern Basin (NB) 253690 (2) 80 / 130

Mixed Conifer,

Ponderosa Pine,

Riparian

Juniper, Aspen, Pinyon-

Juniper, Ponderosa

Pine, Mountain

Mahogany

304 9.7 None

Table 4.2. CLM4 future simulations for control, baseline, and bioenergy management scenarios. Transient CO

2

, nitrogen deposition (Ndep), and land cover files are annual files covering the period from 2010-2100. Ndep and land cover are spatially interpolated using the CLM toolkit from a half-degree global dataset while the CO

2

file is a single value for the entire region which changes annually according to the IPCC representative concentration pathway

RCP4.5.

Name Description

BAU Business-as-Usual

NO_HARV No harvest

Climate

2010-2100

2010-2100

CO

2

/ Ndep Land

Cover

Baseline rcp 4.5 rcp 4.5 rcp 4.5

2000

Harvest

Rate (%)

BAU

None

Insect

Mortality

Yes

Yes

Fire

Yes

Yes

THIN

SALV

CC

TS

TSC

Thin at risk forests

Salvage dead

Clearcut secondary mesic forests

Thin + salvage

Thin + salvage + clearcut

Bioenergy Management

2010-2100

2010-2100

2010-2100

2010-2100

2010-2100 rcp 4.5 rcp 4.5 rcp 4.5 rcp 4.5 rcp 4.5 rcp 4.5 rcp 4.5 rcp 4.5 rcp 4.5 rcp 4.5

50

90

100

50, 90

50, 90, 100

No

Yes

No

Yes

Yes

Yes

Yes

Yes

Yes

Yes

2010-2100

2010-2100

Control

CLIM Vary climate

CLIM_CN Vary climate, CO

2

,

Ndep

* constant level is based on the value from the year 2000 constant* constant rcp 4.5** constant none none

No

No

No

No

** RCP4.5 refers to the IPPCC representative concentration pathway where CO

2

rises to 550ppm by the end of the century

Table 4.3. Ecoregion forested hectares percent of total study region forested area, mean stand ages by ownership, dominant forest types, mean annual precipitation (MAP), mean annual temperature (MAT), and proposed bioenergy management scenarios.

Simulated

Management

Strategy

Description

No Harvest BAU harvest is stopped. This strategy allows for determination of the potential carbon uptake and storage in the absence of harvest.

Salvage Single harvest of biomass from an area susceptible to insect-related mortality. This scenario is hypothetical and intended to test the difference between allowing insect-killed trees to decompose on the land versus using the biomass for products and energy.

Thin Three 30 year thinning rotations. Areas with a mean fire return interval of < 40 years and a portion of the West Cascades were selected. This scenario is intended to test the effectiveness of fire emissions reduction while producing a continuous supply of bioenergy biomass.

Clearcut

Thin +

Salvage

Thin +

Clearcut +

Salvage

Two 45 year clearcut rotations from productive ecoregions with historical clearcut history and forestland capable of 10

Mg of merchantable wood per hectare per year

Application of both Thin and Salvage scenarios above

Application of both Thin and Clearcut scenarios above

Proposed treatable area

(hectares)

12,800,000

596,026

4578689

2642941

5174715

7817656

Forest area treated (%)

0

0.7 to 1.4 % per year

1.5 to 1.7 % per year

0.1 to 0.8 % per year

1 to 2% per year

1 to 2% per year

Harvest Rate

(%), Actual

Removal (%)

0

90%, ~65

50%, ~35%

95%, ~85

50%, ~35

90%, ~65

50%, ~35

95%, ~85

90%, ~65

Restrictions none

Non-reserved forestland, no roadless area

Non-reserved forestland, no roadless area, current stand densities must be

> 300 trees per hectare

Non-reserved, no roadless areas

Same as above for both

Same as above for all three

Table 4.4. Business-as-Usual (BAU) regional carbon stocks and fluxes and soil mineralized nitrogen (Smin) summed over the treatment period (90 years) and the difference compared to BAU for each of the management scenarios. Positive numbers indicate increases over BAU values and negative numbers indicate decreases. A ‘±’ symbol indicates estimate uncertainty where available.

Scenario

Wood (Tg C) Soil (Tg C)

Carbon stocks

CWD

(Tg C) Scenario

BAU (BAU)

Climate (CLIM)

Climate + CN (CLMCN)

Salvage (SALV)

Thin (THIN)

Clearcut (CC)

1436

73

336

-1

-214

-174

Thin + Salvage (TS) -215

Thin + Salvage + Clearcut (TSC) -436

No harvest (NOHARV) 336

NPP

(Tg C)

998

2

10

-1

-14

-13

-15

-5

10

AgNPP

(TgC)

199

22

20

0

-19

-21

-20

-24

20

Carbon fluxes

R h

(Tg C)

-2

-7

0

0

0

-2

0

Leaf C

(Tg C)

55.5

-7

NEP

(Tg C)

-1

-1

1

1

0

0

-1

Lit C

(Tg C)

19.8

-1

Fire

(Tg C)

Tot C

(Tg C)

2709

89

367

-2

-249

-208

-252

-473

367

Harvest

(Tg C)

Smin

(Tg N)

0.078

0.058

0.078

0.078

0.083

0.085

0.083

0.090

0.078

NBP

(Tg C)

BAU (BAU)

Climate (CLIM)

Climate + CN (CLMCN)

Salvage (SALV)

Thin (THIN)

Clearcut (CC)

Thin + Salvage (TS) -62

Thin + Salvage + Clearcut (TSC) 5

No harvest (NOHARV) -62

5568

-578

-66

2

-67

26

3504

-381

-39

4

-41

19

-38

-1

-42

4311

-170

20

-13

-102

-72

-129

-137

47

1257 ± 138

-409

-86

15

35

98

67

142

-109

264

30

30

4

-14

-19

-14

-37

25

524

-472

-472

48

208

339

267

550

-472

469 ± 103

33

356

-37

-160

-212

-187

-372

390

89

Figure 4.1. Study region area (Oregon, USA) divided by ecoregions and coded according to potential management plans (BM = Blue Mountains, CR= Coast Range, CP =

Columbia Plateau, EC = East Cascades, KM = Klamath Mountains, NB = Northern

Basin, SR = Snake River, WC = West Cascades, WV = Willamette Valley). Blue areas indicate no treatment change to BAU, green areas indicate thinning, orange areas are salvaged, and red areas are clearcut harvest. Old-growth reserves were excluded from any proposed harvest in the analysis.

90

Figure 4.2. Predicted regional change in: (A) annual temperature and precipitation from

2010 to 2100, (B) NPP (dashed lines) and R h

(solid lines), (C) NEP, and (D) mineralized soil nitrogen for each of the factors tested (climate, CO

2

and N deposition, BAU harvest).

The x-axis is years and the y-axis is in T g C yr

-1

for NPP, R h

, and NEP and Mg N for mineralized N content. Black line = climate only effects, red line = climate + CO

2

+ N deposition effects, and the green line = BAU.

91

Figure 4.3. The change (delta) for each management scenario compared to BAU in the state of Oregon for a) NPP, b) Rh, c) NEP, d) NBP, e) Total Carbon, and f) Harvest. The x-axis is in years and the y-axis is in Tg C yr

-1

.Orange lines = Salvage harvest, Red lines

= clearcut harvest, Green lines = thinning harvest, Blue lines = thinning + salvage harvest, and Yellow lines = No Harvest. There were no grid cells in the Blue Mountains under the clearcut scenario and no significant changes to BAU in the No Harvest scenario

(excluded from figure). The black line indicates the ‘zero’ change. Positive values indicate increases to BAU carbon stocks and fluxes while negative values indicate decreases.

Figure 4.4. Oregon total net ecosystem production (NEP), fire emissions, wood harvest removals, and net biome production (NBP) in Tg C per year. Each management scenario is included and all scenarios account for future environmental change.

92

93

Figure 4.5: The change (delta) for each management scenario compared to BAU in the

Blue Mountains for a) NPP, b) Rh, c) NEP, d) NBP, e) Total Carbon, and f) Harvest. The x-axis is in years and the y-axis is in Tg C yr

-1

. Orange lines = Salvage harvest, Green lines = thinning harvest, and blue lines = thinning + salvage harvest. There were no grid cells in the Blue Mountains under the clearcut scenario and no significant changes to

BAU in the No Harvest scenario (excluded from figure). The black line indicates the

‘zero’ change. Positive values indicate increases to BAU carbon stocks and fluxes while negative values indicate decreases.

94

Figure 4.6: The change (delta) for each management scenario compared to BAU in the

Coast Range for a) NPP, b) Rh, c) NEP, d)NBP, e) Total Carbon, and f) Harvest. The xaxis is in years and the y-axis is in Tg C yr

-1

. Red lines = Clearcut harvest, Yellow lines =

No harvest. There were no grid cells in the Coast Range under the thinning and salvage scenarios. The black line indicates the ‘zero’ change. Positive values indicate increases to

BAU carbon stocks and fluxes while negative values indicate decreases.

95

Figure 4.7: The change (delta) for each management scenario compared to BAU in the

East Cascades for a) NPP, b) Rh, c) NEP, d) NBP, e) Total Carbon, and f) Harvest. The x-axis is in years and the y-axis is in Tg C yr

-1

. Orange lines = Salvage harvest, Green lines = thinning harvest, Light Blue = thinning + salvage and Yellow lines = no harvest.

There were no grid cells in the East Cascades under the clearcut scenario and no significant changes to BAU in the No Harvest scenario. The black line indicates the

‘zero’ change. Positive values indicate increases to BAU carbon stocks and fluxes while negative values indicate decreases.

96

Figure 4.8: The change (delta) for management scenario compared to BAU in the

Klamath Mountains for a) NPP, b) Rh, c) NEP, d) NBP, e) Total Carbon, and f) Harvest.

The x-axis is in years and the y-axis is in Tg C yr

-1

. Red lines = Clearcut harvest, Yellow lines = No harvest. There were no grid cells in the Coast Range under the thinning and salvage scenarios. The black line indicates the ‘zero’ change. Positive values indicate increases to BAU carbon stocks and fluxes while negative values indicate decreases.

97

Figure 4.9: The change (delta) for each management scenario compared to BAU in the

West Cascades for a) NPP, b) Rh, c) NEP, d)NBP, e) Total Carbon, and f) Harvest. The x-axis is in years and the y-axis is in Tg C yr

-1

. Re lines = Clearcut harvest, Yellow lines

= No harvest and Green lines = thinning harvest. There were no grid cells in the West

Cascades for the salvage scenario. The black line indicates the ‘zero’ change. Positive values indicate increases to BAU carbon stocks and fluxes while negative values indicate decreases.

98

Figure 4.10. Total change in Net C e

to the atmosphere summed over the 90 year period compared to BAU for each management scenario in each ecoregion in Tg C. Error bars indicate the estimate uncertainty (± 5 Tg C) based on the uncertainty in current regional

Net C e

.

99

Figure 4.11. Blue Mountains LCA by component for a) the Thinning Scenario only and b) All scenarios Net C e

compared to BAU (Delta Net C e

). All units are in Tg C yr

-1

.

100

Figure 4.12. Coast Range LCA by component for a) the Clearcut Scenario only and b)

All scenarios Net C e

compared to BAU (Delta Net C e

). All units are in Tg C yr

-1

.

101

Figure 4.13. East Cascade LCA by component for a) the Thinning Scenario only and b)

All scenarios Net C e

compared to BAU (Delta Net C e

). All units are in Tg C yr

-1

.

102

Figure 4.14. Klamath Mountains LCA by component for a) the BAU Scenario only and b) All scenarios Net C e

compared to BAU (Delta Net C e

). All units are in Tg C yr

-1

.

103

Figure 4.15. West Cascade LCA by component for a) the Thinning Scenario only and b)

All scenarios Net C e

compared to BAU (Delta Net C e

). All units are in Tg C yr

-1

.

Chapter 5: Conclusions

Mitigation of climate change with forests requires a thorough understanding of forest carbon uptake and storage and the ability to make predictions about how uptake

104 and storage will change as climate and management change. Policy aimed at reducing greenhouse gases (GHGs), particularly carbon dioxide (CO

2

), through the use of forest biomass as bioenergy must consider the land-based forest carbon dynamics ( in-situ carbon) as well as the off-site wood product sinks and sources. Because climate, natural disturbance, and management affect the rate at which forests store and release CO

2

to the atmosphere (Law et al. 2004, Williams et al. 2012), careful consideration of all of the factors involved is necessary in order to make predictions about net reductions or increases to emissions following proposed bioenergy management strategies. The strength in our predictive ability lies in our ability to reproduce current observed patterns with models. Through a combined approach using observational datasets and terrestrial ecosystem process models, this study analyzed the impacts of proposed forest management for bioenergy on short- and long-term regional carbon dynamics in the US west coast states.

Observations with forest inventory and supplemental plot indicated that the US

West Coast forests are strong carbon sinks under business-as-usual and current environmental conditions (Chapter 2). Current uptake after accounting for removals

(NBP) is 26±3 Tg C yr

-1

(76±9 g C m

2

yr

-1

) for Washington, Oregon, and California combined. This is equivalent to 18% of the total fossil fuel emissions for the region.

Forests cover approximately 50% of the land area on the West Coast.

Using the observed data, we calculated the net carbon emissions to the atmosphere over the next 20 years assuming constant environmental conditions following implementation of low to high intensity bioenergy management. We found that forest thinning for fire prevention measures and large-scale bioenergy harvest in US West Coast forests leads to 2-14% higher emissions compared to current management practices over the next 20 years (Chapter 2; Hudiburg et al. 2011).. However, this was not true for all ecoregions within the study area. In ecoregions where the current sink has been weakened

105 by insect-related mortality, increased fire emissions, or reduced net primary production, management schemes including bioenergy production may succeed in jointly reducing fire risk and carbon emissions. Hence, forest policy should not be a one-size-fits-all approach.

While the short-term implications of bioenergy are important for policy timelines which call for emissions reductions in the next 20 years, we recognize that changing environmental conditions will affect the current forest carbon sink and therefore the net carbon emissions over longer timeframes. Terrestrial ecosystem models provide the means to integrate changing environmental conditions, disturbance, and management on carbon dynamics and make predictions about future carbon dynamics as climate and management change. The uncertainty in the predictions depends on the model’s ability to duplicate historical and current conditions. We evaluated the NCAR C ommunity Land

Model (CLM4), the land model portion of the Community Earth System Model

(CESM1.0.4) in the region (Lawrence et al. 2011) with existing datasets and identified model deficiencies (Chapter 3).

CLM4 was evaluated against inventory data in PNW forests using the default configuration and regional parameterization. Initial results indicated overall underestimation of stem biomass except in the semi-arid Blue Mountains where it was overestimated by 48 Mg C m

-2

. Modeled ecoregion mean NPP was also overestimated in the Blue Mountains and mesic West Cascades. Following initial default parameterization, model improvements were made to account for ecoregion differences in the physiological variables, foliar N content and mortality, and this resulted in an overall improvement in NPP estimates (all ecoregions fell within the observed range of uncertainty). Changing the stem wood allocation algorithm further improved the results, however wood carbon stocks were still underestimated in the West Cascades and Klamath Mountains (CLM4_stem).

After modifications, model estimates of NPP fell within the observed range of uncertainty in all ecoregions and the underestimation of stem biomass was reduced. At 12 km spatial resolution, reduced statistics indicated adequate model performance in all ecoregions for both stem biomass and NPP (reduced was equal 3 and 2, respectively).

106

This was an improvement from the default configuration by 50% for stem biomass and

30% for NPP. Within ecoregions, there was good performance ( < 2) in the Coast

Range and West Cascades, ecoregions that account for ~40% of the forested area. There was generally better performance in mature stands and apart from the Klamath

Mountains, better performance in wet stands.

At the tower sites, modeled annual GPP fell within the observed range of uncertainty (reduced x

2

statistic < 1.2) at both sites. Fall, winter, and spring values followed seasonal patterns, however, summer GPP was consistently underestimated and did not fall within the observed range of uncertainty. Modeled LUE was too low suggesting better estimation of seasonal V cmax

would improve summer GPP estimates.

We found improved parameterization of foliar nitrogen content and nitrogen availability increased monthly GPP in all months and therefore did not improve seasonal dynamics.

This indicates that the model code controlling V cmax

response to either temperature and / or soil water content should be revised.

Following evaluation, a series of modeling experiments were performed to determine the long term impacts of varying management strategies on net carbon emissions compared to BAU (Chapter 4). Simulations of future conditions in Oregon indicate no change to the current sink strength as long as BAU management is maintained. At the end of the 21 st

century, predicted regional net ecosystem production

(BAU NEP) did not change compared with the current NEP indicating no significant influence by changing climate, nitrogen deposition, and increasing CO

2

concentrations over the next 90 years. In other words, statewide rates of net carbon uptake under business-as-usual conditions were not reduced compared to current conditions. However, simulated BAU NBP did decline by ~2 Tg C yr

-1

by the end of the century compared to the beginning primarily due to increases in fire emissions.

For the state of Oregon, implementation of bioenergy management strategies which increase harvest compared to BAU (thinning and clearcut scenarios), did not result in net reductions to carbon emissions over the next 90 years despite inclusion of all wood product and bioenergy substitutions. The management associated increases in net carbon uptake (NEP), reduction in fire emissions, and associated wood product sinks and

substitutions are not sufficiently large enough to balance the life cycle emissions from

107 product acquisition, conversion, decomposition, and combustion for energy by the end of the 21 st

century.

There were some cases where the dominant trend at the regional level varied at different spatial resolutions. In the semi-arid ecoregions (Blue Mountains and East

Cascades), net carbon uptake following thinning began to improve over BAU at the end of the 90 year period suggesting bioenergy management could lead to net reductions over a longer time period (> 100 years). For salvage logging, there was no change in emissions compared to BAU by the end of the 21 st

century in all cases. Finally, in the Klamath

Mountains, there was evidence that the current level of harvest, wood products, and bioenergy management reduces emissions compared to management plans with no harvest at all.

Management and policy implications

Regional bioenergy management strategies utilizing forest biomass will not reduce GHGs, particularly carbon dioxide, over the next 90 years given the results of this study. The specific situations where emissions were not increased (i.e. Salvage logging) are limited to single harvests that do not increase BAU harvest by more than 4%. In the majority of the ecoregions and at the state-level, policy that reduces harvest would decrease emissions over the 90 year period. This is particularly important for policy makers as this reemphasizes that a ‘one-size-fits-all’ approach will not help to mitigate climate change. Management plans will need to be written accounting for local conditions, especially where predicted future local climate deviates from regional climate

(i.e. precipitation in the Blue Mountains). Bioenergy from forest biomass by thinning in the Blue Mountains was beginning to reduce emissions compared to BAU at the end of the 21s century. Based on this analysis, the challenge for policy makers and land managers will be to balance wood harvest for local economies and forest health while maintaining carbon emissions relative to BAU management as significant reductions do not seem possible.

108

Suggestions for future research

The short- and long-term experiments could still be improved to determine the net carbon emissions from bioenergy management strategies that will provide a continuous bioenergy supply (i.e. rotation based thinning and clearcut harvest). Longer simulations and better fire and insect related mortality prediction will be required to identify areas where long term bioenergy management strategies will be most effective. Also, the simulations do not include a mechanism for PFTs to change overtime (i.e. dynamic vegetation model) due to stress caused by disturbance factors such as insects, disease and fire or drought. The annual scenario harvest rates presented here are not necessarily reflective of industry operations, which would require a steady supply of wood over space and time. The temporal and spatial distribution of the harvests will ultimately need to be planned with mill and facility location and capacity in mind. Finally, more climate and RCP scenarios should be investigated to provide a range of responses to policy makers and land managers.

References

Hudiburg, T. W., B. E. Law, C. Wirth, and S. Luyssaert. 2011. Regional carbon dioxide implications of forest bioenergy production. Nature Clim. Change 1:419-423.

Law, B. E., D. Turner, J. Campbell, O. J. Sun, S. Van Tuyl, W. D. Ritts, and W. B.

Cohen. 2004. Disturbance and climate effects on carbon stocks and fluxes across

Western Oregon USA. Global Change Biology 10:1429-1444.

Lawrence, D., K. W. Oleson, M. G. Flanner, P. E. Thorton, S. C. Swenson, P. J.

Lawrence, X. Zeng, Z.-L. Yang, S. Levis, K. Skaguchi, G. B. Bonan, and A. G.

Slater. 2011.

Parameterization Improvements and Functional and Structural Advances in Version 4 of the Community Land Model. J. Adv. Model. Earth Syst. 3:27.

Siemers, E. 2011. Northwest Energy plans $300M Oregon biomass investment. Pages http://www.sustainablebusinessoregon.com/articles/2011/2003/northwest-energyplans-2300m-oregon.html in Portland Business Journal, Portland.

109

Williams, C. A., G. J. Collatz, J. Masek, and S. N. Goward. 2012. Carbon consequences of forest disturbance and recovery across the conterminous United States. Global

Biogeochem. Cycles 26:GB1005.

Bibliography

Battles, J., T. Robards, A. Das, K. Waring, J. Gilless, G. Biging, and F. Schurr. 2008.

Climate change impacts on forest growth and tree mortality: a data-driven modeling study in the mixed-conifer forest of the Sierra Nevada, California.

Climatic Change 87:193-213.

Birdsey, R. A. 2007. North American forests in The First State of the Carbon Cycle

Report (SOCCR): The North American Carbon Budget and Implications for the

Global Carbon Cycle. U.S. Climate Change Science Program and the

Subcommittee on Global Change Research, Washington, DC.

Bonan, G. B. 2008. Forests and Climate Change: Forcings, Feedbacks, and the Climate

Benefits of Forests. Science 320:1444-1449.

Bonan, G. B., P. J. Lawrence, K. W. Oleson, S. Levis, M. Jung, M. Reichstein, D. M.

Lawrence, and S. C. Swenson. 2011. Improving canopy processes in the

Community Land Model version 4 (CLM4) using global flux fields empirically inferred from FLUXNET data. J. Geophys. Res. 116:G02014.

Bonan, G. B., K. W. Oleson, R. A. Fisher, G. Lasslop, and M. Reichstein. 2012.

Reconciling leaf physiological traits and canopy flux data: Use of the TRY and

FLUXNET databases in the Community Land Model version 4. J. Geophys. Res.

117:G02026.

Breshears, D. D., O. B. Myers, C. W. Meyer, F. J. Barnes, C. B. Zou, C. D. Allen, N. G.

McDowell, and W. T. Pockman. 2008. Tree die-off in response to global changetype drought: mortality insights from a decade of plant water potential measurements. Frontiers in Ecology and the Environment 7:185-189.

Buyx, A., and J. Tait. 2011. Ethical Framework for Biofuels. Science.

Campbell, J., D. Donato, D. Azuma, and B. Law. 2007. Pyrogenic carbon emission from a large wildfire in Oregon, United States. J. Geophys. Res. 112.

Campbell, J., G. Alberti, J. Martin, and B. E. Law. 2009. Carbon dynamics of a ponderosa pine plantation following a thinning treatment in the northern Sierra

Nevada. Forest Ecology and Management 257:453-463.

Campbell, J. L., M. E. Harmon, and S. R. Mitchell. 2012. Can fuel-reduction treatments really increase forest carbon storage in the western US by reducing future fire emissions? Frontiers in Ecology and the Environment 10:83-90.

110

111

Coops, N. C., and R. H. Waring. 2011. Estimating the vulnerability of fifteen tree species under changing climate in Northwest North America. Ecological Modelling

222:2119-2129.

Curtis, P. S., and X. Z. Wang. 1998. A meta-analysis of elevated CO2 effects on woody plant mass, form and physiology. Oecologia 113:299-313.

DeLucia, E. H., D. J. Moore, and R. J. Norby. 2005. Contrasting responses of forest ecosystems to rising atmospheric CO2: Implications for the global C cycle.

Global Biogeochemical Cycles 19:doi:10.1029/2004GB002346.

Edburg, S. L., J. A. Hicke, D. M. Lawrence, and P. E. Thornton. 2011. Simulating coupled carbon and nitrogen dynamics following mountain pine beetle outbreaks in the western United States. J. Geophys. Res. 116:G04033.

Eidenshink, J. et al. A project for monitoring trends in burn severity. Fire ecology 3, 3-20

(2007).

Evangelista, P. H., S. Kumar, T. J. Stohlgren, and N. E. Young. 2011. Assessing forest vulnerability and the potential distribution of pine beetles under current and future climate scenarios in the Interior West of the US. Forest Ecology and Management

262:307-316.

Evans, A., and A. Finkral. 2009. From renewable energy to fire risk reduction: a synthesis of biomass harvesting and utilization case studies in US forests. GCB

Bioenergy 1:211-219.

Fargione, J., J. Hill, D. Tilman, S. Polasky, and P. Hawthorne. 2008. Land clearing and the biofuel carbon debt. Science 319:1235-1238.

Farquhar, G. D., S. v. Caemmerer, and J. Berry. 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 149:78-90.

French, N. H. F., W. J. de Groot, L. K. Jenkins, B. M. Rogers, E. Alvarado, B. Amiro, B. de Jong, S. Goetz, E. Hoy, E. Hyer, R. Keane, B. E. Law, D. McKenzie, S. G.

McNulty, R. Ottmar, D. R. Pérez-Salicrup, J. Randerson, K. M. Robertson, and

M. Turetsky. 2011. Model comparisons for estimating carbon emissions from

North American wildland fire. J. Geophys. Res. 116:G00K05.

Göckede, M., A. M. Michalak, D. Vickers, D. P. Turner, and B. E. Law. 2010.

Atmospheric inverse modeling to constrain regional-scale CO2 budgets at high spatial and temporal resolution. J. Geophys. Res. 115:D15113.

112

Groenendijk, M., A. J. Dolman, C. Ammann, A. Arneth, A. Cescatti, D. Dragoni, J. H. C.

Gash, D. Gianelle, B. Gioli, G. Kiely, A. Knohl, B. E. Law, M. Lund, B.

Marcolla, M. K. van der Molen, L. Montagnani, E. Moors, A. D. Richardson, O.

Roupsard, H. Verbeeck, and G. Wohlfahrt. 2012a. Seasonal variation of photosynthetic model parameters and leaf area index from global Fluxnet eddy covariance data. J. Geophys. Res. 116:G04027.

Groenendijk, M., A. J. Dolman, M. K. van der Molen, R. Leuning, A. Arneth, N.

Delpierre, J. H. C. Gash, A. Lindroth, A. D. Richardson, H. Verbeeck, and G.

Wohlfahrt. 2012b. Assessing parameter variability in a photosynthesis model within and between plant functional types using global Fluxnet eddy covariance data. Agricultural and Forest Meteorology 151:22-38.

Gurney, K. R., D. L. Mendoza, Y. Zhou, M. L. Fischer, C. C. Miller, S. Geethakumar, and S. de la Rue du Can. 2009. High Resolution Fossil Fuel Combustion CO2

Emission Fluxes for the United States. Environmental Science & Technology

43:5535-5541.

Gustavsson, L., P. l. Börjesson, B. Johansson, and P. Svenningsson. 1995. Reducing

CO2 emissions by substituting biomass for fossil fuels. Energy 20:1097-1113.

Harmon, M. E., W. K. Ferrell, and J. F. Franklin. 1990. Effects on carbon storage of conversion of old-growth forests to young forests. Science 247:699-702.

Harmon, M. E., J. M. Harmon, W. K. Ferrell, and D. Brooks. 1996. Modeling carbon stores in Oregon and Washington forest products: 1900–1992. Climatic Change

33:521-550.

Harmon, M. E., and B. Marks. 2002. Effects of silvicultural practices on carbon stores in

Douglas-fir – western hemlock forests in the Pacific Northwest, U.S.A.: results from a simulation model. Canadian Journal of Forest Research/Revue Canadienne de Recherche Forestiere 32:863-877.

Hayes, D. J., D. P. Turner, G. Stinson, A. D. McGuire, Y. Wei, T. O. West, L. S. Heath,

B. de Jong, B. G. McConkey, R. A. Birdsey, W. A. Kurz, A. R. Jacobson, D. N.

Huntzinger, Y. Pan, W. M. Post, and R. B. Cook. 2012. Reconciling estimates of the contemporary North American carbon balance among terrestrial biosphere models, atmospheric inversions, and a new approach for estimating net ecosystem exchange from inventory-based data. Global Change Biology 18:1282-1299.

Heath, L. S., V. Maltby, R. Miner, K. E. Skog, J. E. Smith, J. Unwin, and B. Upton.

2010. Greenhouse Gas and Carbon Profile of the U.S. Forest Products Industry

Value Chain. Environmental Science & Technology 44:3999-4005.

Hicke, J. A., J. C. Jenkins, D. S. Ojima, and M. Ducey. 2008. Spatial patterns of forest

113 characteristics in the western United States derived from inventories. Ecological

Applications.

Hudiburg, T., B. E. Law, D. P. Turner, J. Campbell, D. Donato, and M. Duane. 2009.

Carbon dynamics of Oregon and Northern California forests and potential landbased carbon storage. Ecological Applications 19:163-180.

Hudiburg, T. W., B. E. Law, C. Wirth, and S. Luyssaert. 2011. Regional carbon dioxide implications of forest bioenergy production. Nature Clim. Change 1:419-423.

Huggett Jr, R. J., K. L. Abt, and W. Shepperd. 2008. Efficacy of mechanical fuel treatments for reducing wildfire hazard. Forest Policy and Economics 10:408-414.

Hurteau, M. D., and M. North. 2010. Carbon recovery rates following different wildfire risk mitigation treatments. Forest Ecology and Management 260:930-937.

Hurtt, G. C., S. Frolking, M. G. Fearon, B. Moore, E. Shevliakova, S. Malyshev, S. W.

Pacala, and R. A. Houghton. 2006. The underpinnings of land-use history: three centuries of global gridded land-use transitions, wood-harvest activity, and resulting secondary lands. Global Change Biology 12:1208-1229.

ICCT. 2010. Carbon intensity of crude oil in Europe. International Council on Clean

Transportation, Washington D.C.

IPCC. 2007. IPCC expert meeting report: Technical Summary, Towards new scenarios for analysis of emissions, climate change, impacts, and response strategies.

Intergovernmental Panel on Climate Change, Cambridge.

Janssens, I. A., W. Dieleman, S. Luyssaert, J. A. Subke, M. Reichstein, R. Ceulemans, P.

Ciais, A. J. Dolman, J. Grace, G. Matteucci, D. Papale, S. L. Piao, E. D. Schulze,

J. Tang, and B. E. Law. 2010. Reduction of forest soil respiration in response to nitrogen deposition. Nature Geosci 3:315-322.

Jenkins, J. C., D. C. Chojnacky, L. S. Heath, and R. A. Birdsey. 2003. National-Scale

Biomass Estimators for United States Tree Species. Forest Science 49:12-35.

Jenkinson, D. S., K. Goulding, and D. S. Powlson. 1999. Nitrogen deposition and sequestration. Nature 400:629.

Keith, H., B. G. Mackey, and D. B. Lindenmayer. 2009. Re-evaluation of forest biomass carbon stocks and lessons from the world's most carbon-dense forests.

Proceedings of the National Academy of Sciences, USA 106:11635-11640.

Kolb, T. E., J. K. Agee, P. Z. Fulé, N. G. McDowell, K. Pearson, A. Sala, and R. H.

114

Waring. 2007. Perpetuating old ponderosa pine. Forest Ecology and Management

249:141-157.

Krishnan, P., T. A. Black, R. S. Jassal, B. Chen, and Z. Nesic. 2009. Interannual variability of the carbon balance of three different-aged Douglas-fir stands in the

Pacific Northwest. J. Geophys. Res. 114:G04011.

Kurz, W. A., G. Stinson, G. J. Rampley, C. C. Dymond, and E. T. Neilson. 2008. Risk of natural disturbances makes future contribution of Canada's forests to the global carbon cycle highly uncertain. Proceedings of the National Academy of Sciences

105:1551-1555.

Law, B. E., and M. E. Harmon. 2011. Forest sector carbon management, measurement and verification, and discussion of policy related to climate change. Carbon

Management 2:73-84.

Law, B., T. Hudiburg, and S. Luyssaert. 2012. Thinning Effects on Forest Productivity:

Consequences of Preserving Old Forests and Mitigating Impacts of Fire and Drought.

Plant Ecology & Diversity doi: 10.1080/17550874.2012.679013

Law, B. E., O. J. Sun, J. Campbell, S. Van Tuyl, and P. E. Thornton. 2003. Changes in carbon storage and fluxes in a chronosequence of ponderosa pine. Global Change

Biology 9:510-524.

Law, B. E., D. Turner, J. Campbell, O. J. Sun, S. Van Tuyl, W. D. Ritts, and W. B.

Cohen. 2004. Disturbance and climate effects on carbon stocks and fluxes across

Western Oregon USA. Global Change Biology 10:1429-1444.

Lawrence, D., K. W. Oleson, M. G. Flanner, P. E. Thorton, S. C. Swenson, P. J.

Lawrence, X. Zeng, Z.-L. Yang, S. Levis, K. Skaguchi, G. B. Bonan, and A. G.

Slater. 2011. Parameterization Improvements and Functional and Structural

Advances in Version 4 of the Community Land Model. J. Adv. Model. Earth Syst.

3:27.

Lippke, B., E. Oneil, R. Harrison, K. Skog, L. Gustavsson, and R. Sathre. 2011. Life cycle impacts of forest management and wood utilization on carbon mitigation: knowns and unknowns. Carbon Management 2:303-333.

Lippke, B., J. Wilson, J. Meil, and A. Taylor. 2010. Characterizing the importance of carbon stored in wood products. Wood and Fiber Science 42:5-14.

Luyssaert, S., P. Ciais, S. L. Piao, E.-D. Schulze, M. Jung, S. Zaehle, M. J. Schelhaas, M.

Reichstein, G. Churkina, D. Papale, G. Abril, C. Beer, J. Grace, D. Loustau, G.

Matteucci, F. Magnani, G. J. Nabuurs, H. Verbeeck, M. Sulkava, G. R. v. d.

WERF, and I. A. Janssens. 2009. The European carbon balance: part 3: Forests.

115

Global Change Biology 16:1429-1450.

Luyssaert, S., E. D. Schulze, A. Borner, A. Knohl, D. Hessenmoller, B. E. Law, P. Ciais, and J. Grace. 2008. Old-growth forests as global carbon sinks. Nature 455:213-

215.

Malmsheimer, R. W., J. L. Bowyer, J. S. Fried, E. Gee, R. L. Izlar, R. A. Miner, I. A.

Munn, E. Oneil, and W. C. Stewart. 2011. Managing Forests because Carbon

Matters: Integrating Energy, Products, and Land Management Policy. Journal of

Forestry 109:S7-S51.

Marland, G., and B. Schlamadinger. 1997. Forests for carbon sequestration or fossil fuel substitution? A sensitivity analysis. Biomass and Bioenergy 13:389-397.

Matson, P., L. Johnson, C. Billow, J. Miller, and R. Pu. 1994. Seasonal Patterns and

Remote Spectral Estimation of Canopy Chemistry Across the Oregon Transect.

Ecological Applications 4:280-298.

McDowell, N., J. R. BROOKS, S. A. FITZGERALD, and B. J. BOND. 2003. Carbon isotope discrimination and growth response of old Pinus ponderosa trees to stand density reductions. Plant, Cell and Environment 26:631-644.

McDowell, N. G., C. D. Allen, and L. Marshall. 2010. Growth, carbon-isotope discrimination, and drought-associated mortality across a Pinus ponderosa elevational transect. Global Change Biology 16:399-415.

McKechnie, J., S. Colombo, J. Chen, W. Mabee, and H. L. MacLean. 2011. Forest

Bioenergy or Forest Carbon? Assessing Trade-Offs in Greenhouse Gas Mitigation with Wood-Based Fuels. Environmental Science & Technology 45:789-795.

Meigs, G., D. Donato, J. Campbell, J. Martin, and B. Law. 2009. Forest fire impacts on carbon uptake, storage, and emission: the role of burn severity in the Eastern

Cascades, Oregon. Ecosystems 12:1246-1267.

Mitchell, S. R., M. E. Harmon, and K. E. B. O'Connel. 2009. Forest fuel reduction alters fire severity and long-term carbon storage in three Pacific Northwest ecosystems.

Ecological Applications 19.

Mitchell, S. R., M. E. Harmon, and K. E. B. O'Connell. 2012. Carbon debt and carbon sequestration parity in forest bioenergy production. GCB Bioenergy:n/a-n/a.

116

Morris, D. M. 2009. Changes in DOC and DON fluxes in response to harvest intensity of black-spruce-dominated forest ecosystems in northwestern Ontario. Canadian

Journal of Soil Science 89:67-79.

Morris, J. 2010. Bury or Burn North America MSW? LCAs Provide Answers for Climate

Impacts & Carbon Neutral Power Potential. Environmental Science &

Technology 44:7944-7949.

Moss, R. H., J. A. Edmonds, K. A. Hibbard, M. R. Manning, S. K. Rose, D. P. van

Vuuren, T. R. Carter, S. Emori, M. Kainuma, T. Kram, G. A. Meehl, J. F. B.

Mitchell, N. Nakicenovic, K. Riahi, S. J. Smith, R. J. Stouffer, A. M. Thomson, J.

P. Weyant, and T. J. Wilbanks. 2010. The next generation of scenarios for climate change research and assessment. Nature 463:747-756.

Mote, P. W., and E. Salathé. 2010. Scenarios: Future Climate in the Pacific Northwest.

University of Washington, Seattle.

NRC. 2010a. Advancing the Science of Climate Change. National Academy of Sciences,

National Academies Press, Washington D.C.

NRC. 2010b. Verifying Greenhouse Gas Emissions National Academies Press,

Washington, D.C.

Nunery, J. S., and W. S. Keeton. 2010. Forest carbon storage in the northeastern United

States: Net effects of harvesting frequency, post-harvest retention, and wood products. Forest Ecology and Management 259:1363-1375.

Oleson, K. W., D. M. Lawrence, G. B. Bonan, M. G. Flanner, E. Kluzek, P. J. Lawrence,

S. Levis, S. C. Swenson, P. E.Thornton, A. Dai, M. Decker, R. DIckinson, J.

Feddema, C. L. Heald, F. Hoffman, J. F. Lamarque, N. Mahowald, G.-Y. Niu, T.

Qian, J. Randerson, S. Running, K.Sakaguchi, A. Slater, R. Stockli, A. Wang, Z.-

L. Yang, X. Zeng, and X. Zeng. 2010. Technical description of version 4.0 of the

Community Land Model. NCAR Tech. Note NCAR/TN-478+STR, 257.

Omernik, J. M. 1987. Ecoregions of the conterminous United States. Map (scale

1:7,500,000). Annals of the Association of American Geographers 77:118-125.

Ottmar, R. D., S. J. Pritchard, R. E. Vihnanek, and D. V. Sandberg. 2006. Modification and Validation of Fuel Consumption Models for Shrub and Forested lands in the

Southwest, Pacific Northwest, Rockies,Midwest, Southeast and Alaska, Final

Report JFSP Project 98-1-9-06.

Pacala, S. W. 2010. Verifying Greenhouse Gas Emissions: Methods to Support

International Climate Agreements. NRC The National Academies Press,

Washington, D.C.

Qian, T., A. Dai, K. E. Trenberth, and K. W. Oleson. 2006. Simulation of Global Land

Surface Conditions from 1948 to 2004. Part I: Forcing Data and Evaluations.

Journal of Hydrometeorology 7:953-975.

Raffa, K. F., B. H. Aukema, B. J. Bentz, A. L. Carroll, J. A. Hicke, M. G. Turner, and W.

117

H. Romme. 2008. Cross-scale Drivers of Natural Disturbances Prone to

Anthropogenic Amplification: The Dynamics of Bark Beetle Eruptions.

Bioscience 58:501-517.

Raymond, C. L., and D. L. Peterson. 2005. Fuel treatments alter the effects of wildfire in a mixed-evergreen forest, Oregon, USA. Canadian Journal of Forest Research

35:2981-2995.

Richter Jr, D., D. H. Jenkins, J. T. Karakash, J. Knight, L. R. McCreery, and K. P.

Nemestothy. 2009. Resource policy: wood energy in America. Science 323:1432-

1433.

Rogers, B., R. Neilson, R. Drapek, J. Lenihan, J. Wells, D. Bachelet, and B. Law. 2011.

Impacts of climate change on fire regimes and carbon stocks of the U.S. Pacific

Northwest. J. Geophys. Res. 116 G03037.

Routa, J., S. KellomÄKi, A. KilpelÄInen, H. Peltola, and H. Strandman. 2011. Effects of forest management on the carbon dioxide emissions of wood energy in integrated production of timber and energy biomass. GCB Bioenergy 3:483-497.

Ryan, M. G. 2010. Temperature and tree growth. Tree Physiology 30:667-668.

Salathé, E., L. Leung, Y. Qian, and Y. Zhang. 2010. Regional climate model projections for the State of Washington. Climatic Change 102:51-75.

Salathe, E. P., P. W. Mote, and M. W. Wiley. 2007. Review of scenario selection and downscaling methods for the assessment of climate change impacts on hydrology in the United States pacific northwest. International Journal of Climatology

27:1611-1621.

Schaefer, K. M., C. R. Schwalm, C. A. Williams, M. Arain, A. G. Barr, J. M. Chen, and

K. J. Davis. In Press. A model-data comparison of Gross Primary Productivity:

Results from the North American Carbon Program site synthesis. Journal of

Geophysical Research.

118

Schwalm, C. R., C. A. Williams, K. Schaefer, R. Anderson, M. A. Arain, I. Baker, A.

Barr, T. A. Black, G. Chen, J. M. Chen, P. Ciais, K. J. Davis, A. Desai, M. Dietze,

D. Dragoni, M. L. Fischer, L. B. Flanagan, R. Grant, L. Gu, D. Hollinger, R. C.

Izaurralde, C. Kucharik, P. Lafleur, B. E. Law, L. Li, Z. Li, S. Liu, E. Lokupitiya,

Y. Luo, S. Ma, H. Margolis, R. Matamala, H. McCaughey, R. K. Monson, W. C.

Oechel, C. Peng, B. Poulter, D. T. Price, D. M. Riciutto, W. Riley, A. K. Sahoo,

M. Sprintsin, J. Sun, H. Tian, C. Tonitto, H. Verbeeck, and S. B. Verma. 2010. A model-data intercomparison of CO2 exchange across North America: Results from the North American Carbon Program site synthesis. J. Geophys. Res.

115:G00H05.

Schulze, E. D., W. D. Vries, M. Hauhs, K. Rosén, L. Rasmussen, O.-C. Tann, and J.

Nilsson. 1989. Critical loads for nitrogen deposition in forest ecosystems. Water,

Air, and Soil Pollution 48:451-456.

Schulze, E.-D., C. Körner, B. E. Law, H. Haberl, and S. Luyssaert. 2012. Large-scale bioenergy from additional harvest of forest biomass is neither sustainable nor greenhouse gas neutral. GCB Bioenergy:n/a-n/a.

Searchinger, T. D., S. P. Hamburg, J. Melillo, W. Chameides, P. Havlik, D. M. Kammen,

G. E. Likens, R. N. Lubowski, M. Obersteiner, M. Oppenheimer, G. Philip

Robertson, W. H. Schlesinger, and G. David Tilman. 2009. Fixing a critical climate accounting error. Science 326:527-528.

Siemers, E. 2011. Northwest Energy plans $300M Oregon biomass investment. Pages http://www.sustainablebusinessoregon.com/articles/2011/2003/northwest-energyplans-2300m-oregon.html in Portland Business Journal, Portland.

Smith, J. E., L. Heath, K. E. Skog, and R. Birdsey. 2006. Methods for Calculating Forest

Ecosystem and Harvested Carbon with Standard Estimates for Forest Types of the

United States Gen. Tech. Rep. NE-343. U.S. Department of Agriculture, Forest

Service, Northeastern Research Station, Newtown Square, PA.

Solomon, S., G.-K. Plattner, R. Knutti, and P. Friedlingstein. 2009. Irreversible climate change due to carbon dioxide emissions. Proceedings of the National Academy of

Sciences.

Sonne, E. 2006. Greenhouse Gas Emissions from Forestry Operations. J. Environ. Qual.

35:1439-1450.

Stephens, S. L., J. J. Moghaddas, C. Edminster, C. E. Fiedler, S. Haase, M. Harrington, J.

E. Keeley, E. E. Knapp, J. D. McIver, K. Metlen, C. N. Skinner, and A.

Youngblood. 2009. Fire treatment effects on vegetation structure, fuels, and potential fire severity in western US forests. Ecological Applications 19:305-320.

119

Stinson, G., W. A. Kurz, C. E. Smyth, E. T. Neilson, C. C. Dymond, J. M. Metsaranta, C.

Boisvenue, G. J. Rampley, Q. Li, T. M. White, and D. Blain. 2011. An inventorybased analysis of Canada's managed forest carbon dynamics, 1990 to 2008.

Global Change Biology 17:2227-2244.

Thomas, C. K., B. E. Law, J. Irvine, J. G. Martin, J. C. Pettijohn, and K. J. Davis. 2009.

Seasonal hydrology explains interannual and seasonal variation in carbon and water exchange in a semiarid mature ponderosa pine forest in central Oregon. J.

Geophys. Res. 114:G04006.

Thornton, P. E., B. E. Law, H. L. Gholz, K. L. Clark, E. Falge, D. S. Ellsworth, A. H.

Goldstein, R. K. Monson, D. Hollinger, and M. Falk. 2002. Modeling and measuring the effects of disturbance history and climate on carbon and water budgets in evergreen needleleaf forests. Agricultural and Forest Meteorology

113:185-222.

Thornton, P. E., S. W. Running, and M. A. White. 1997. Generating surfaces of daily meteorological variables over large regions of complex terrain. J. Hydrol.

190:214–251.

Thornton, P. E., and N. E. Zimmermann. 2007. An improved canopy integration scheme for a land surface model with prognostic canopy structure. Journal of Cllimate

20:3902-3923.

Thornton, P. E., J.-F. o. Lamarque, N. A. Rosenbloom, and N. M. Mahowald. 2007.

Influence of carbon-nitrogen cycle coupling on land model response to CO2 fertilization and climate variability. Global Biogeochem. Cycles 21.

Thornton, P. E., and N. E. Zimmermann. 2007. An improved canopy integration scheme for a land surface model with prognostic canopy structure. Journal of Cllimate

20:3902-3923.

Turner, D. P., D. Ritts, B. E. Law, W. Cohen, Z. Yang, T. Hudiburg, J. L. Campbell, and

M. Duane. 2007. Scaling net ecosystem production and net biome production over a heterogeneous region in the western United States. Biogeosciences 4:597-612.

Turner, D. P., D. Ritts, B. E. Law, W. Cohen, Z. Yang, T. Hudiburg, J. L. Campbell, and

M. Duane. 2007. Scaling net ecosystem production and net biome production over a heterogeneous region in the western United States. Biogeosciences 4:597-612.

Upton, B., R. Miner, M. Spinney, and L. S. Heath. 2008. The greenhouse gas and energy impacts of using wood instead of alternatives in residential construction in the

United States. Biomass and Bioenergy 32:1-10.

120

USGS. 2009. The National Map LANDFIRE: LANDFIRE National Existing Vegetation

Type layer. Pages http://gisdata.usgs.net/website/landfire/ in U.S. Department of

Interior, Geological Survey. U.S. Department of Interior, Geological Survey. van Mantgem, P. J., N. L. Stephenson, J. C. Byrne, L. D. Daniels, J. F. Franklin, P. Z.

Fule, M. E. Harmon, A. J. Larson, J. M. Smith, A. H. Taylor, and T. T. Veblen.

2009. Widespread Increase of Tree Mortality Rates in the Western United States.

Science 323:521-524. van der Werf, G. R., J. T. Randerson, L. Giglio, G. J. Collatz, M. Mu, P. S. Kasibhatla, D.

C. Morton, R. S. DeFries, Y. Jin, and T. T. van Leeuwen. 2010. Global fire emissions and the contribution of deforestation, savanna, forest, agricultural, and peat fires (1997–2009). Atmos. Chem. Phys. 10:11707-11735.

Van Tuyl, S., B. E. Law, D. P. Turner, and A. I. Gitelman. 2005. Variability in net primary production and carbon storage in biomass across Oregon forests--an assessment integrating data from forest inventories, intensive sites, and remote sensing. Forest Ecology and Management 209:273-291.

Walker T, Cardellichio P, C. A, and E. al. 2010. Biomass sustainability and carbon policy study. NCI-2010-03 (2010). Manomet Center for Conservation Sciences, ,

Massachusetts.

Waring, R. H., and J. F. Franklin. 1979. Evergreen Coniferous Forests of the Pacific

Northwest. Science 204:1380-1386.

Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. Warming and

Earlier Spring Increase Western U.S. Forest Wildfire Activity. Science 313:940-

943.

Whitaker, J., K. E. Ludley, R. Rowe, G. Taylor, and D. C. Howard. 2010. Sources of variability in greenhouse gas and energy balances for biofuel production: a systematic review. GCB Bioenergy 2:99-112.

White, M. A., P. E. Thornton, S. W. Running, and R. R. Nemani. 2000. Parameterization and Sensitivity Analysis of the BIOMEBGC Terrestrial Ecosystem Model: Net

Primary Production Controls. Earth Interactions 4:1-85.

Wiedinmyer, C., and M. D. Hurteau. 2010. Prescribed fire as a means of reducing forest carbon emissions in the western United States. Environmental Science &

Technology 44:1926-1932.

121

Williams, C. A., G. J. Collatz, J. Masek, and S. N. Goward. 2012. Carbon consequences of forest disturbance and recovery across the conterminous United States. Global

Biogeochem. Cycles 26:GB1005.

Winford, E., and J. Gaither Jr. In Press. Carbon outcomes from fuels treatment and bioenergy production in a Sierra Nevada forest Forest Ecology and Management.

Wirth, C., K. Ogle, A. Nuske, J. T. Weedon, W. K. Cornwell, and J. Cornelissen. 2010.

Global analysis of wood decomposition – the roles of environment, taxonomy, and traits. in. Max Planc Institute for Biogeochemistry.

Wright, L., B. Boundy, R. Perlack, S. C. Davis, and B. Saulsbury. 2006. Biomass energy data book: edition 1, 1 edition. Oak Ridge National Laboratory, Oak Ridge,

Tennessee, USA.

Zhang, F., J. M. Chen, Y. Pan, R. A. Birdsey, S. Shen, W. Ju, and L. He. 2012.

Attributing carbon changes in conterminous U.S. forests to disturbance and nondisturbance factors from 1901 to 2010. J. Geophys. Res. 117:G02021.

APPENDICES

122

Appendix A: Supporting information for Chapters 2 and 4

Methods

General Approach: We combined data from the Forest Inventory and Analysis

123

National Program (FIA) for Oregon, Washington and California (~34 million hectares) with LandFire satellite remote sensing products (USGS 2009), 200 Appendix plots

(Hudiburg et al. 2009), and a global wood decomposition database(Wirth et al. 2010) to provide new estimates of US West Coast forest biomass carbon stocks, net ecosystem production (NEP, the net of photosynthesis and respiration), net biome production (NBP, the net forest carbon-sink accounting for removals) and their uncertainties. We include all forestland in our analysis. These forests range across all age classes (0-1000 years old), are on rotation management or may have never been harvested, and are both public and privately owned. Plot values were aggregated by climatic region (ecoregion), age class

(succession class), and forest type and the mean values were used to assign a value to each associated 30 meter pixel. Thinning treatments were applied to each plot according to a specific set of criteria (Figure A1b) of removed forest carbon. Life cycle assessment

(LCA) was calculated at the ecoregion, state, and regional level using published values for associated fossil fuel emissions, energy conversion efficiencies, wood product pool ratios and decomposition, and fossil fuel inputs (Table A4).

Database: Federal forest inventory data (FIA) are now being collected on an annual basis, statewide on all types of forestland in all regions. The inventory design consists of 0.404 hectare (one-acre) plots systematically placed across the landscape, encompassing a representative range of stand ages, disturbance histories, ownerships, and land cover types. The study area includes all forested land in Washington, Oregon and

California for the period from 2001-2006. In addition to the traditional tree surveys, new measurements include woody detritus, understory shrubs, and litter allowing for a more complete quantification of land-based carbon stocks (excluding stocks in soils). As of

2006, there were 8889 measured plots (Table A7) with accessible forestland (Plot Status code = 2), of which 8659 with tree increment data recorded which is required for our methods of estimating NPP. Of these 8659, only 6840 had the necessary detritus

measurements to calculate NEP (Phase 3 plots). These remaining 6840 plots are still

124 distributed across the landscape in all forest types and ecoregions with about 20% of the plots excluded. Plot means of biomass, NPP, NEP, and NBP were scaled to state totals using spatially explicit forest cover, forest type and productivity, and succession class data products available in 30 x 30 meter resolution from LandFire Landsat satellite remote sensing derived products (USGS 2009). Finally, state carbon budgets were estimated using datasets containing annual harvest removals and wood densities, fire emissions, and fossil fuel emissions (Table 5).

Biomass stock and flux estimates: Tree and shrub carbon stocks and NPP were calculated with a combination of species-specific allometric equations, tree increment data, and supplemental plot data using methods (Table A8) described in (Hudiburg et al.

2009). Wood cores collected from a subsample of trees on each plot were used to estimate plot NPP from 10 year diameter growth increment and thus the annual mean

NPP is averaged over a 10 year period of growth conditions. Woody detritus carbon was calculated using the line intersect method and species- and decay class-specific wood densities (Hudiburg et al. 2009). Foliage litter and duff depth measurements were converted to biomass as the product of the depth and the material density. Foliage NPP was calculated by dividing foliage biomass per tree by the average foliage retention time

(average number of years of foliage that a stand carries). Herb mass was estimated using a biomass conversion factor and percent cover on each plot.

We define NEP as the difference between annual NPP and heterotrophic respiration. While direct measurements of soil respiration are not available on FIA plots, we were able to calculate NEP using a mass-balance approach and Appendix plot data

(Giardina and Ryan 2002, Campbell et al. 2009):

NEP = Aboveground NPP – dead wood decomposition – litterfall + Δ root +Δ soil

C. (1)

The basic assumption is that annual soil respiration is balanced with litterfall, belowground carbon allocation and change in carbon of the roots and soils. Aboveground

NPP from tree increment cores and dead wood species, diameter, and decay class are available from West Coast FIA observations. Deadwood decomposition was estimated

125 based on a global dataset of wood decomposition rates of tree species (http://www.bgcjena.mpg.de/bgc-organisms/pmwiki.php/Research/FET). Predictive models used genusspecific baseline rates modified by coefficients describing the sensitivity to mean annual temperature, annual precipitation sum, stem diameter and position (downed versus standing). Litterfall was estimated as foliage NPP minus an average mass retention of 21 percent (Table A9). We assumed change in soil and fine root carbon were zero for plots older than 150 years (old growth) and used the difference between values found at our supplemental plots for younger and mature age classes and the old growth plots to calculate a delta soil carbon (Law et al. 2003, Campbell et al. 2009).

Net Biome Production (NBP): We followed (Chapin et al. 2006) and defined net biome production (NBP) as NEP minus any losses due to fire or harvest. Average annual state timber harvest volumes provided by the respective state Departments of Forestry were converted to biomass removals using known wood densities (Turner et al. 2007).

Actual fire emissions were calculated using burn area and severity (Eidenshink et al.

2007) and biomass specific combustion factors for the region (Table A9). Estimates of fire emissions vary greatly depending on the approach (Ottmar et al. 2006, Campbell et al. 2007, Meigs et al. 2009, Wiedinmyer and Hurteau 2010) and combustion factors used.

We use biomass specific combustion factors from studies in our region (Campbell et al.

2007, Meigs et al. 2009) which include several of the dominant forest types in our region covering 67% of the forested area. For woody detritus, these factors range from 3 – 100% combustion depending on fire severity and the type of biomass. Emissions estimates using these factors compare well with modeled estimates which also distinguish between biomass components (Ottmar et al. 2006). Other studies use a single combustion factor

(30%) for all above ground woody biomass (Wiedinmyer et al. 2006, Wiedinmyer and

Hurteau 2010) including the standing dead trees. Since the majority of the woody biomass is in standing tree boles, this results in an overestimation of fire emissions.

Nevertheless, we have included additional estimates of NBP for both current and treatment emissions using the single combustion factor to determine the sensitivity of this parameter on our results (Table A6).

Treatment Design: To test the effect of biomass thinning on land-based carbon

126 stocks and NBP, we chose three different management scenarios designed to meet varying objectives (Figure A1a). Basal area removals, maximum tree bole size, and areas treated were varied by the following strategies: 1) Fire Prevention: Thinning targets smaller trees (Stephens et al. 2009a) and is limited to areas with frequent fires or short mean fire return intervals (the latter being derived from LandFire (USGS 2009) Landsat remote sensing derived data product). This scenario is unlikely to be economically feasible due to low value of the extracted biomass. 2) Economically Feasible: Thinning targets larger trees followed by smaller trees providing at least 9 Mg of dry biomass per hectare (4.5 Mg C ha

-1

) of merchantable biomass (Skog et al. 2008). Merchantable biomass would help pay for removing fuel ladders (understory trees that allow fire to access and ignite the canopy). Hence, this treatment is limited to areas with short mean fire return interval. 3) Bioenergy Production: Thinning targets all regions and trees to maximize biomass available for energy production (Perlack et al. 2005). This scenario is also expected to be economically feasible because merchantable wood is removed. 4)

Business-as-usual (BAU): forest management remains the same as current practices (14.7

Tg C harvested annually) with no additional thinning or harvest treatments. The biomass removal targets were defined by current or proposed practices (USDA 2010). Treatments were all designed to produce stands capable of resisting crown fire by reducing canopy bulk density and removal of understory ladder fuel (Stephens et al. 2009a). Synthesis of existing fuel reduction treatment studies found that stand basal area was reduced by an average of 48% (Evans and Finkral 2009). We chose to use a range of basal area reductions, 30% for the Fire Prevention treatment, 40% for Economically Feasible, and

50% for the Bioenergy treatment based on current and proposed practices (Finkral and

Evans 2008, Skog et al. 2008, Harrod et al. 2009, Stephens et al. 2009a, Stephens et al.

2009b). These practices are designed to follow a general standard which is to alter stand conditions so that projected fire severity would result in at least 80% of the dominant and codominant residual trees surviving a wildfire under the 80 th

percentile fire weather conditions known as the ”80-80” rule (Stephens et al. 2009a).

127

The main difference between the Fire Prevention and the Economic Production scenario is the basal area reduction allowed and the size of the trees removed. For

Economic Production, larger trees are targeted first and then smaller trees are removed to meet the basal area reduction requirements. We did not intend to predict or analyze the economic cost of any of the treatments and simply allowed for higher DBH removals to help offset the cost. There have been several studies that investigate the economic potential of forest bioenergy (OFRI 2006, Skog et al. 2008, Evans and Finkral 2009), with mixed results (see http://www.eenews.net/climatewire/2011/07/22/5 ). In the Fire

Prevention scenario, trees are removed from smaller DBH classes first and then larger trees are removed until basil area reduction requirements are met. The main difference between the Economic and the Bioenergy scenario is the basal area reduction and the land area treated. Bioenergy Production treats all land regardless of fire risk or return interval while the Economic scenario only treats land with a fire return interval of less than 40 years.

Thinning prescriptions: Thinning prescriptions were applied to the FIA plots for each of the three treatments (Fire Prevention, Economically Feasible, and Bioenergy

Production) according to the management scenarios. The primary objective in each scenario was to reduce stand density in order to reduce the risk of wildfire. The actual fuel reduction thinning treatments applied across different forest types and ownerships vary from stand to stand and are therefore prescribed on a stand-by-stand basis. In order to prevent wildfire, removal of ladder fuels and reduction in crown density are necessary

(Agee and Skinner 2005, Huggett Jr et al. 2008, Harrod et al. 2009). Stand prescriptions usually involve a vegetation model simulation which takes inputs of stand characteristics such as height, species composition, understory structure, canopy bulk density, ground fuels, wind speed, temperature, and moisture conditions. While most of the necessary inputs are available for a given FIA plot at a given time, some of these conditions change over time (weather) and are too stand specific (structure) to be extrapolated to other stands spatially and temporally. We chose to use the average basal area reduction (30-

50%) found in a synthesis of studies (Skog et al. 2008, Evans and Finkral 2009)to insure adequate reduction in crown density and prevent removal of too much biomass.

128

To simulate effective removal of ladder fuels, FIA plots were treated by removing the understory non-merchantable small trees (< 12.7 cm) in all three scenarios. Thinning of overstory trees was varied by scenario. All trees on the plot were grouped into small, medium, and large DBH classes. In the Fire Prevention scenario, the majority of the trees removed were in the small DBH class and less trees were removed from the medium and large classes to a maximum of 30% basal area reduction. An upper DBH limit of 45 cm was set. In the Economically Feasible scenario, the majority of trees were removed from the medium and large DBH class followed by a smaller percentage in the smaller class to a maximum of 40% basal area reduction. In the Bioenergy Production scenario, trees were removed similar to the Fire Prevention scenario but the maximum basal area reduction was 50%. For both the Economically Feasible and the Bioenergy Production scenarios, the DBH limit was set to 60 cm. This upper limit on DBH is currently stated in active forest policy in Oregon and California (ORSenate 2005) and the smaller one was proposed to retain larger trees (Harrod et al. 2009). Bole, branch and bark biomass were considered ‘removed’ from the site and separated into merchantable and bioenergy pools.

Total plot removals were aggregated and mapped by ecoregion, forest type, and age class.

State and regional totals include only non-reserved, productive forestland in accessible areas ( http://www.fs.fed.us/r1/gis/thematic_data/ira_us_dd.htm

). Productive forestland must be capable of producing 10 Mg/ha of merchantable wood annually.

All scenarios exclude public forest reserves and remove all non-merchantable wood (diameter at breast height or DBH < 12.7 cm). A treatment period of 20 years was assumed to be the amount of time required to treat the entire landscape or 5% of the treatable forested area per year (Bowyer 2006). We chose the 40-year mean fire return interval because a plot that is at least half-way through a 40 year mean fire return interval could burn during a 20-year treatment period. FIA plots on forestland capable of producing 10 Mg of merchantable wood per hectare per year were thinned according to each treatment and new plot mean biomass values were scaled to state and ecoregion boundaries to determine the removal totals. The treatment removals were treated as additional harvest (harvest in addition to Business-as-Usual harvest) in further

calculations accounting for the portion of biomass utilized for energy and the portion

129 used as merchantable biomass.

Treatments were assumed to be 75% effective at reducing fire emissions

(Raymond and Peterson 2005). Because only 5% of the landscape is treated each year, associated reductions in fire emissions increase as more forestland is treated. If treatments are 75% effective, then emissions are reduced by 3.75% each year with 75% reduction in the final treatment year. This results in reducing fire emissions by half when integrated over the whole treatment period. We also assume there is no increase in fire frequency (or probability) over the study period due to other factors such as climate change. Fire frequency and intensity are expected to increase in the western US due to climate change (Westerling et al. 2006, Littell et al. 2009), but the extent is highly unknown and limited by the capability of the climate-fire models.

Post-treatment NEP: While increased growth of the remaining trees following thinning is well documented, stand-level NPP is reduced (Law of Constant Final Yield,

(Shinozaki and Kira 1956)) because ultimately resources limit growth not density

(Campbell et al. 2009). Thinning effects on NEP are not well documented and response is variable (Campbell et al. 2009). However, we needed to account for regrowth, either due to the increased growth of remaining trees or the growth of the understory over the treatment period. Since we could not estimate regrowth NEP using the plot data, we decided to use the NEP associated with young-aged plots since this might most mimic the conditions following release from competition. The thinned plots were assigned an NEP equal to the observed mean NEP of stands aged between 1 and 20 years over the treatment period resulting in an overall higher NEP for thinned stands, biasing our results towards beneficial effects on the carbon-sink for forest treatments.

Given the large uncertainty associated with predicting future fire (Rogers et al.,

2011), we assume that in BAU, fire will occur with variability that has been observed in the past (no increases in fire). Since FIA plots include those that have within fire perimeters, post-fire NEP is part of the current flux estimates. However, because we assume no increase in fire, we do not predict a new post-fire NEP, which would beyond the appropriate use of the data. In a recent study by (Raffa et al. 2008), post-fire NEP for

130 the 5 years prior to the Oregon Biscuit fire averaged the same as the five years after the fire. In this study, we have shown that a doubling of fire emissions by using different combustion coefficients still does not compensate for the emissions associated with bioenergy production (Supplemental discussion of sensitivity analysis).

Life-cycle assessment (Figure A1b; Tables 4 and 5): Life-cycle assessment of forest carbon removals includes forestry related sinks and sources of carbon to and from the atmosphere and the associated impact on total fossil fuel emissions (FFE). The C emissions to the atmosphere for each scenario (FCO

2

) over 20 years were calculated as the difference between the sources and the sinks following this process:

FCO2 = NBP + Total Harvest – WD1 – WD2 – Wood Industry FFE – Bioenergy

Emissions + Bioenergy Substitution + FF Well-To-Tank Emissions displacement+

(Wood Substitution) (2)

Where, WD1 is the wood lost during manufacturing processes, WD2 is the wood decomposed over time from product use and wood substitution is included with the assumption that there is an increased demand for wood supply.

To quantify the change in FCO

2

for each scenario we calculate the difference between each scenario and the BAU FCO

2

emissions. The physical sinks are forest net uptake (NBP) and wood products (Harvest) and the added virtual sinks of bioenergy and wood product substitution (FF Substitution). Because the benefits of wood substitution require an increase in wood use and this saturates quickly, we calculate the change in

CO

2

with and without a wood substitution benefit (Law and Harmon 2011). We exclude imports and exports from the study region since we are only interested in quantifying domestic wood production emissions and exports are less than 1% of harvested merchantable wood (http://www.fs.fed.us/pnw/ppet/). ‘Emissions’ include release of carbon from woody biomass combustion, and FFE associated with harvest (Sonne 2006), transport of both harvested material and end-use products (Evans and Finkral 2009,

Heath et al. 2010), and processing and manufacturing of wood products (Heath et al.

2010) and bioenergy (Jaeger et al. 2007, Whitaker et al. 2010). ‘Decomposition’ includes loss of material through decomposition or combustion during the manufacturing of wood

131 products (Smith et al. 2006), and the percentage of wood products that are expected to no longer be in-use at the end of the treatment period (Smith et al. 2006).

Biomass utilized for wood products can end up in a long term storage product

(structural wood) or a short term product (paper). Some wood product carbon reenters the atmosphere through rapid (paper) or slow (wood) decomposition or combustion while some is eventually disposed in landfills where it is very slowly decomposed. West Coast harvests generate merchantable bole wood at rates of 50-60% of the total wood harvested

(Harmon et al. 1996) and an average of 54% of this wood remains in use or is in landfills after 20 years (Smith et al. 2006). Using tables provided by (Smith et al. 2006) we determined the amounts of long and short term wood products that could be generated by the merchantable wood harvested accounting for the losses along the way and multiplied this by 54% to determine the wood product storage (Smith et al. 2006). Because this ratio could increase or decrease due to changes in manufacturing efficiency, product use, or recycling, we allowed for a 10% increase and decrease of this percentage for the additional harvest only as part of the sensitivity analysis (Heath et al. 2010, Lippke et al.

2010). The remaining non-merchantable wood (including understory trees) from harvest was used for biofuel biomass and associated emissions.

Fossil fuel substitution with bioenergy was calculated as a 50/50 energy mix of ethanol conversion and biomass combustion compared to fossil fuel derived automotive gasoline. Woody biomass provides less energy per unit of carbon emitted than fossil fuels

(i.e. wood has an energy content of 20 GJ per ton versus 43.5 GJ per ton in automotive gasoline because fossil fuels have a lower heating value (Wright et al. 2006)). Under maximum yield conditions, the potential energy of woody biomass is 78% of fossil fuel if combusted and 36% if converted to cellulosic ethanol (Mitchell et al. 2009). These maximum values are highly unrealistic as they have yet to be obtained (Mitchell et al.

2009) and ratios up to 30% lower have been suggested (Galbe and Zacchi 2002).

Nevertheless, we use the maximum values in our estimates under optimum conditions and reduce the ratios by 10 and 20% to provide a range of conditions in the sensitivity analysis (Table 7). State annual fossil fuel emissions were acquired from the Vulcan

Project Database (http://www.purdue.edu/eas/carbon/vulcan/index.php) and from the

Oregon Department of Energy

(http://www.oregon.gov/ENERGY/GBLWRM/CCIG.shtml, Appendix A).

There are also emissions associated with crude extraction and manufacturing, sometimes called the wells-to-tank emissions (WTT). Fossil fuel LCA total emissions

132

(wells to wheels; WTW) include both WTT and tank-to-wheels (TTW) emissions. The amount of C emitted per unit of fossil fuel energy varies widely by oil field, but generally

WTT emissions are about 15%(ICCT 2010) of total emissions (WTW), or 12 g CO

2

per

MJ of energy. We have included theses emissions in the Wood Industry FFE and we have added a WTT displacement benefit along with the bioenergy substitution benefit.

An additional estimate of the LCA was calculated for a wood product substitution benefit. Wood product substitution for a 50/50 mix of aluminum and steel used in residential American housing generates a 36% reduction in fossil fuel emissions assuming a maintained rate in new residential housing (Upton et al. 2008). We applied a wood substitution benefit as 36% of the final structural wood product pool.

Uncertainty Analysis: Monte Carlo simulations were used to conduct an uncertainty analysis using the mean and standard deviations for NPP and Rh calculated using several approaches. For NPP, three alternative sets of allometric equations were used to estimate the uncertainty due to variation in region and/or species-specific allometry. The full suite of species-specific equations that use tree diameter (DBH) and height (preferred) were compared to a DBH-only national set (Jenkins et al. 2003) and to a grouped forest type set. For, Rh, the variation in the calculated decomposition rate was used to quantify the uncertainty. A species-specific lookup table of decay constants was compared to decay constants that were allowed to vary by genus, precipitation, and temperature or by class, precipitation and temperature. Finally, uncertainty in NBP was calculated as the combined uncertainty of NEP, fire emissions (10%) (Campbell et al.

2007), harvest emissions (7%) (Heath and Smith 2000), and land cover estimates (10%) using the propagation of error approach (NRC 2010). Uncertainty estimates are represented in the Figures by the grey error bars and in tables with ‘±’.

Sensitivity analysis of most of the LCA parameters is summarized below (Table

A6). The most sensitive parameters in this study that affect net emissions and NBP are

133 land area treated, allowable removals (DBH limit and basal area reductions) per unit area, and to some extent, fire emissions. First, we present a range of scenarios that vary by land area and allowable removals. Removals are varied by basal area reduction limits of 30,

40, and 50% for the FP, EC, and BP scenarios respectively. These reductions equate to removal rates of 25-53% of live biomass. To test the sensitivity of the reduction in fire emissions, we also calculated NBP and net emissions assuming 50% and 75% effectiveness of the treatments. Additional parameters that affect only net emissions is the ratio of wood products that are in use at the end of the treatment period, the efficiency of the conversion to energy, and the fossil fuel inputs required for energy conversion. For the in-use product ratio, we calculated net emissions for a 10% increase and decrease to the ratio. For the conversion efficiencies, we varied each of these factors by 100, 90 and

80% of the maximum possible values to present a range of results reflecting the most optimum conditions (100% efficiency) to the least optimum (80% efficiency) (Table A6).

Values for the most optimum conditions are represented in the Figures unless otherwise noted.

In addition to evaluating the sensitivity of the proposed treatments to the parameter estimates, we also explored the effect of varying the range of harvest to NEP ratios, wood product to bioenergy biomass ratios, percent combusted versus converted to cellulosic ethanol, fossil fuel inputs required, and amount of wood product in the shortterm product pool with subsequent recapture as bioenergy. We determined the hypothetical ratios where the forest net carbon flux was zero (neutral) or greater than or equal to the current flux (BAU) and compared the net carbon flux for each scenario with the range of ratios (Figure A6).

Discussion

State-level Estimates: Forest carbon stocks (excluding soil carbon) for the entire region are 5.0 ± 0.8 Pg C with 31% in Washington, 36% in Oregon, and 33% in

California (Table A2). NPP ranges from 100 to 900 g C m

-2

across the region and falls within the range of 100 to 1600 g C m

-2

yr

-1

reported for temperate and boreal forests

(Luyssaert et al. 2008) and Rh ranges from 100 to 600 g C m

-2

. Our estimates are in line

with previous work: Our mean NEP ranges from -50 to 400 g C m

-2

yr

-1

similar to the

134 range of -50 to 800 g C m

-2

yr

-1

reported for temperate forests (Luyssaert et al. 2008).

Using a simulation model, the total NEP of Oregon in the late 1990’s was estimated to be

17 ± 11 Tg C yr

-1

(vs. 15.3 ± 1.6 Tg C yr

-1

in this study) most of which was attributed to forests (Turner et al. 2007). Furthermore, recent estimates from (Raffa et al. 2008) predict an NEP of 25.5 Tg C yr

-1

(vs. 29.2 Tg C yr

-1

for the same area in this study) in the northwest forest plan area of Washington, Oregon, and California. Also using a simulation model, the total NEP of California (Potter 2010) for 2001-2004 ranged from

14-24 Tg C yr

-1

(vs.18.1 Tg C yr

-1

in this study). Previous regional estimates of NEP were not found for Washington.

We explored four scenarios, three treatments and business-as-usual (Figure A1a).

The removal limits of 30-50% of stand basal area resulted in 25– 53 % removal of aboveground live tree biomass per plot which is typical for fuel reduction treatments

(Skog et al. 2008, Harrod et al. 2009). These treatments do not replace current management practices. They result in additional harvest above the current harvest in the region. Statewide removals were much lower (by 5-10 Tg C yr

-1

) in Washington than the other two states for the Fire Prevention and Economically Feasible scenarios due to a higher median MFRI (91-100 years versus less than 60 years) resulting in a reduced treatment area (Table A2). The Bioenergy Production scenario results in 264, 220, and

92% reductions in NBP in Washington, Oregon, and California respectively with

Washington and Oregon forests becoming a carbon source (Table A2). The Fire

Prevention and Economically Feasible scenario had the most impact on California NBP compared to Oregon and Washington (decreased from 13.6 Tg C yr

-1

to 5.6 and 9.4 Tg C yr

-1

respectively). Furthermore, at the state level, Washington NBP was not significantly different from BAU for either the Fire Prevention or the Economically Feasible scenario because a smaller percentage of the forested area is in a high fire risk area compared to the other two states resulting in lower harvest levels. Washington removals were balanced by the assumed reduction in fire emissions and increased NEP from regrowth.

These findings suggest that in regions where proposed harvest is low there may be little effect on NBP compared with BAU.

Comparing California and Oregon, the area-weighted state level differences in

135

NEP, fire and harvest, are respectively 20, 3.1 and 30 g C m

-2

y

-1 summing to a difference in NBP of 46.4 g C m

-2

y

-1

(Table A3), indicating the largest differences were in NEP and harvest removals. Coastal Redwood forests in California, for example, contribute 16.5 g

C m

-2

y

-1 more to the state-wide NBP per unit area than the same forest type in Oregon.

However, the opposite was observed in, for example, North Pacific Maritime Mesic-Wet

Douglas-fir-Western Hemlock Forest. This forest type has a 19.9 g C m

-2

y

-1 higher NBP in Oregon than in California. Although there are considerable differences in fire emissions between Oregon and California (1.3 and 1.8 Tg C in OR vs CA), the difference in NEP and harvest between similar forest types within the same ecoregions appears to be the dominant reason of the observed differences in NBP between the two states. Some of the possible causes of the difference in NBP are: (1) A productive forest type is present in

California but absent in Oregon (i.e. Lower Montane Blue Oak-Foothill Pine Woodland and Savanna), (2) A forest type with an above-average NBP is managed similarly in

California and Oregon but is more abundant and productive in California (i.e. Coastal

Redwood Forest and Mediterranean (Dry-)Mesic Mixed Conifer Forest and Woodland) and (3) The losses through harvest are lower in California than Oregon in forest types that cover a large area (North Pacific Sitka Spruce, Dry-Mesic and Wet-Mesic Douglasfir-Western Hemlock Forests). Additionally, our estimates of NBP rely on our estimates of NEP using the mass balance approach. For stands older than 150 years, we have assumed a steady-state for the soil carbon pool (i.e. delta soil carbon is zero). For stands younger than 150 years, we used the best available data to estimate a delta soil carbon value for each plot depending on the age. We may be overestimating by a larger amount if the loss of soil carbon due to disturbance is higher (or gain is slower) in ecoregions (i.e

California) not represented in our soils dataset. If we are overestimating NEP and subsequently overestimating NBP, the biomass removal impact on NBP is underestimated.

Previous estimates of forest biomass potentially available for energy supply in

Oregon and California vary widely, from 0.4 – 14 Tg C yr

-1

, depending on assumptions of area needing treatment, volume removed per hectare, and the number of years over

which treatments are conducted (Bowyer 2006, Williams 2007, Strittholt and Tutak

2009). While these estimates are in general agreement with our estimates from all three treatments (3.8 – 17 Tg C yr

-1

for Oregon and California combined), they only addressed

136 a portion of the potential biomass removals i.e. that used for energy production. In our scenario analysis, we go beyond previous approaches by accounting for the fate of all biomass removals and assess their effect on forest NBP. The thinning treatments result in additional biomass removals of 11- 44 Tg C yr

-1

from current inventory biomass levels for the entire region (Table A5) of which only 7 – 24 Tg C yr

-1

would be used for energy supply. Since current harvest levels are half of what they were in 1980s, three times current harvest rates are possible in this region as has been suggested in government and industry reports and given the current level of effort that is going into developing a biomass industry (Perlack et al. 2005, OFRI 2006). The proposed harvest intensities are simply scenarios that are being considered in one region, and the application of such scenarios over other regions or subregions may not be appropriate given the forest type, climate, and management history. For example, in fire prone or beetle killed areas, it may be necessary to apply moderate harvest levels. Our approach lends itself to testing the carbon consequences of location-specific management activities.

Additional LCA analysis: Proper accounting of the in-situ NBP in LCA reveals the effects of forest management on atmospheric CO

2

when considering mitigation options for reducing CO

2 emissions. Towards this aim, we developed a conceptual model to determine the outcomes of mitigation options which may include different ratios of wood product to energy mixes, higher or lower BAU harvest to NEP ratios, or efficiency of fossil fuel usage. The conceptual LCA model addresses the main determinants of the forest sector carbon budget i.e. forest management and wood processing. This conceptual model reveals that the largest decreases in the forest sector emissions are accomplished by reducing the harvest to NEP ratio (Figure A3, red line).

We analyzed the sink-strength of the forest sector for varying management intensities where intensity was expressed as the ratio of harvest to net ecosystem productivity (H/NEP), the latter including forest growth and regrowth. The current

H/NEP ratio is about 0.3. Wood processing was described by the ratio of wood products

137 to total harvest (P/H). We then investigated the combinations of P/H ratios varying from

0 to 1 with different combinations of bioenergy (combustion versus ethanol), 0-100% reduction in fossil fuel inputs, and recycling of manufactured waste for bioenergy production to determine if biomass end-use affected the forest sector CO

2 emissions.

Changing the ratios for the percentage used in wood products versus bioenergy (Figure

A3a), the mix of energy used for combustion versus cellulosic ethanol (Figure A3b), the efficiency of fossil fuel inputs (Figure A3c), and the reduction of the short-term product pool (Figure A3d) has very little impact compared to the increase in removals.

Furthermore, this analysis suggests that a reduction of net CO

2

emissions compared to BAU can only be realized if harvest remains at current levels or increases to a maximum of 20% more than BAU, but this requires that either all bioenergy is produced by means of combined heat and power rather than ethanol (Figure A3b), or wood-use results in 100% reduction of fossil fuel emissions from this process (Figure

A3c), or 100% of waste wood is used for bioenergy production (Figure A3d). These measures are definitely unlikely to take place within the proposed 20-year time-frame.

Sensitivity Analysis: The differences in NBP and emissions due to land area treated and allowable removals are shown in Table 2. The amount of land area treated has a significant impact on NBP. In the life-cycle assessment, the range in efficiencies changed the impact of the scenarios by 3-28% (Table A6). For example, if the amount of wood products in use are reduced by 10%, bioenergy production is at 80% of optimum conditions (least effective), and fire emissions are reduced by 50%, net emissions to the atmosphere increase by a larger amount: 101, 251, and 579 Tg C for the Fire Prevention,

Economically Feasible, and Bioenergy Production scenarios respectively (compared with

44, 175, and 421 Tg C for 100% of optimum conditions; Table A5). Using the alternative combustion estimates reduces the impact of the FP and EC scenarios by 6-9%, and increases the impact in the BP scenario. This increase in initial fire emissions eliminates the net increases in emissions for the FP scenario (very small annual increase), but only under optimum conditions. The impact is greatest in the Klamath Mountains. Inclusion of wood substitution reduces atmospheric emissions by 2-10% across the scenarios, but only under optimum conditions and assuming there is a demand for the wood use.

Finally, our estimates of BAU harvest practices may decrease in the future, in

138 which case, we could be overestimating removals over the next 20 years. However, this is unlikely because harvest declined significantly since 1990 due to restrictions placed on harvest on federal lands as part of the Northwest Forest Plan.

References

Agee, J. K., and C. N. Skinner. 2005. Basic principles of forest fuel reduction treatments.

Forest Ecology and Management 211:83-96.

Bowyer, J. 2006. Biomass Energy and Biofuels from Oregon's Forests. Oregon Forest

Resources Institute, Corvallis, OR.

Campbell, J., G. Alberti, J. Martin, and B. E. Law. 2009. Carbon dynamics of a ponderosa pine plantation following a thinning treatment in the northern Sierra

Nevada. Forest Ecology and Management 257:453-463.

Campbell, J., D. Donato, D. Azuma, and B. Law. 2007. Pyrogenic carbon emission from a large wildfire in Oregon, United States. J. Geophys. Res. 112.

Campbell, J. L., O. J. Sun, and B. E. Law. 2004a. Supply-side controls on soil respiration among Oregon forests. Global Change Biology 10:1857-1869.

Chapin, F., G. Woodwell, J. Randerson, E. Rastetter, G. Lovett, D. Baldocchi, D. Clark,

M. Harmon, D. Schimel, R. Valentini, C. Wirth, J. Aber, J. Cole, M. Goulden, J.

Harden, M. Heimann, R. Howarth, P. Matson, A. McGuire, J. Melillo, H.

Mooney, J. Neff, R. Houghton, M. Pace, M. Ryan, S. Running, O. Sala, W.

Schlesinger, and E. D. Schulze. 2006. Reconciling carbon-cycle concepts, terminology, and methods. Ecosystems 9:1041-1050.

Eidenshink, J., B. Schwind, K. Brewer, Z. L. Zhu, B. Quayle, S. Howard, and F. USDA.

2007. A project for monitoring trends in burn severity. Fire ecology 3:3-20.

Evans, A., and A. Finkral. 2009. From renewable energy to fire risk reduction: a synthesis of biomass harvesting and utilization case studies in US forests. GCB

Bioenergy 1:211-219.

Finkral, A. J., and A. M. Evans. 2008. The effects of a thinning treatment on carbon stocks in a northern Arizona ponderosa pine forest. Forest Ecology and

Management 255:2743-2750.

139

Galbe, M., and G. Zacchi. 2002. A review of the production of ethanol from softwood.

Applied Microbiology and Biotechnology 59:618-628.

Giardina, C. P., and M. G. Ryan. 2002. Total belowground carbon allocation in a fastgrowing eucalyptus plantation estimated using a carbon balance approach.

Ecosystems 5:487-499.

Harmon, M. E., J. M. Harmon, W. K. Ferrell, and D. Brooks. 1996. Modeling carbon stores in Oregon and Washington forest products: 1900–1992. Climatic Change

33:521-550.

Harmon, M. E., and J. Sexton. 1996. Guidelines for Measurements of Woody Debris in

Forest Ecosystems. . Publication No. 20. US LTER Network Office, University of

Washington, Seattle, WA.

Harrod, R. J., D. W. Peterson, N. A. Povak, and E. K. Dodson. 2009. Thinning and prescribed fire effects on overstory tree and snag structure in dry coniferous forests of the interior Pacific Northwest. Forest Ecology and Management

258:712-721.

Heath, L. S., V. Maltby, R. Miner, K. E. Skog, J. E. Smith, J. Unwin, and B. Upton.

2010. Greenhouse Gas and Carbon Profile of the U.S. Forest Products Industry

Value Chain. Environmental Science & Technology 44:3999-4005.

Heath, L. S., and J. E. Smith. 2000. An assessment of uncertainty in forest carbon budget projections. Environmental Science and Policy 3:73-82.

Hudiburg, T., B. E. Law, D. P. Turner, J. Campbell, D. Donato, and M. Duane. 2009.

Carbon dynamics of Oregon and Northern California forests and potential landbased carbon storage. Ecological Applications 19:163-180.

Huggett Jr, R. J., K. L. Abt, and W. Shepperd. 2008. Efficacy of mechanical fuel treatments for reducing wildfire hazard. Forest Policy and Economics 10:408-414.

ICCT. 2010. Carbon intensity of crude oil in Europe. International Council on Clean

Transportation, Washington D.C.

Jaeger, W. K., R. Cross, and T. M. Egelkraut. 2007. Biofuel Potential in Oregon:

Background and Evaluation of Options. 1078, Oregon State University, Corvallis,

OR.

Jenkins, J. C., D. C. Chojnacky, L. S. Heath, and R. A. Birdsey. 2003. National-Scale

Biomass Estimators for United States Tree Species. Forest Science 49:12-35.

140

Keyes, M. R., and C. C. Grier. 1981. Above- and belowground net production in 40-yearold Douglas-fir stands on low and high productivity sites Canadian Journal of

Forest Research 11:599-605.

Laboratory, F. P. 1974. Wood Handbook: Wood as an Engineering Material Forest

Products Laboratory, Madison, WI.

Law, B. E., and M. E. Harmon. 2011. Forest sector carbon management, measurement and verification, and discussion of policy related to climate change. Carbon

Management 2:73-84.

Law, B. E., O. J. Sun, J. Campbell, S. Van Tuyl, and P. E. Thornton. 2003. Changes in carbon storage and fluxes in a chronosequence of ponderosa pine. Global Change

Biology 9:510-524.

Lippke, B., J. Wilson, J. Meil, and A. Taylor. 2010. Characterizing the importance of carbon stored in wood products. Wood and Fiber Science 42:5-14.

Littell, J. S., D. McKenzie, D. L. Peterson, and A. L. Westerling. 2009. Climate and wildfire area burned in western U.S. ecoprovinces, 1916–2003. Ecological

Applications 19:1003-1021.

Luyssaert, S., E. D. Schulze, A. Borner, A. Knohl, D. Hessenmoller, B. E. Law, P. Ciais, and J. Grace. 2008. Old-growth forests as global carbon sinks. Nature 455:213-

215.

Maeglin, R. R., and H. E. Wahlgren. 1972. Western Wood Density Survey: Report

Number 2. . USDA Forest Service Research Paper FPL-183.

Meigs, G., D. Donato, J. Campbell, J. Martin, and B. Law. 2009. Forest fire impacts on carbon uptake, storage, and emission: the role of burn severity in the Eastern

Cascades, Oregon. Ecosystems 12:1246-1267.

Mitchell, S. R., M. E. Harmon, and K. E. B. O'Connel. 2009. Forest fuel reduction alters fire severity and long-term carbon storage in three Pacific Northwest ecosystems.

Ecological Applications 19.

NRC. 2010. Verifying Greenhouse Gas Emissions National Academies Press,

Washington, D.C.

OFRI. 2006. Oregon Biofuels and Biomass: Woody Biomass in Oregon - Current Uses,

Barriers and Opportunities for Increased Utilization and Research Needs. OSU

Chemical Engineering Department, Institute for Natural Resources, and Oregon

Wood Innovation Center, p.37.

ORSenate. 2005. Oregon Senate Bill 1072. 73rd OREGON LEGISLATIVE

ASSEMBLY--2005 Regular Session, Salem, OR.

Ottmar, R. D., S. J. Pritchard, R. E. Vihnanek, and D. V. Sandberg. 2006. Modification and Validation of Fuel Consumption Models for Shrub and Forested lands in the

Southwest, Pacific Northwest, Rockies,Midwest, Southeast and Alaska, , . Final

Report JFSP Project 98-1-9-06.

141

Perlack, R., L. Wright, A. Turhollow, R. L. Graham, B. Stokes, and D. Erbach. 2005.

Biomass as Feedstock for a Bioenergy and Bioproducts Industry: The Technical

Feasibility of a Billion-Ton Annual Supply. US Department of Energy, US

Department of Agriculture, Oak Ridge, TN.

Potter, C. 2010. The carbon budget of California. Environmental Science & Policy

13:373-383.

Raffa, K. F., B. H. Aukema, B. J. Bentz, A. L. Carroll, J. A. Hicke, M. G. Turner, and W.

H. Romme. 2008. Cross-scale Drivers of Natural Disturbances Prone to

Anthropogenic Amplification: The Dynamics of Bark Beetle Eruptions.

Bioscience 58:501-517.

Raymond, C. L., and D. L. Peterson. 2005. Fuel treatments alter the effects of wildfire in a mixed-evergreen forest, Oregon, USA. Canadian Journal of Forest Research

35:2981-2995.

Shinozaki, K., and T. Kira. 1956. Intraspecific competition among higher plants. VII.

Logistic theory of the C - D effect. J Inst Polytech Osaka City Univ.

Skog, K. E., P. J. Ince, H. Spelter, A. Kramp, and R. J. Barbour. 2008. Woody biomass supply from thinnings to reduce fire hazard in the U.S. West and its potential impact on regional wood markets. Forest Products Society, Madison, WI.

Smith, J. E., L. Heath, K. E. Skog, and R. Birdsey. 2006. Methods for Calculating Forest

Ecosystem and Harvested Carbon with Standard Estimates for Forest Types of the

United States Gen. Tech. Rep. NE-343. U.S. Department of Agriculture, Forest

Service, Northeastern Research Station, Newtown Square, PA.

Sonne, E. 2006. Greenhouse Gas Emissions from Forestry Operations. J. Environ. Qual.

35:1439-1450.

Stephens, S. L., J. J. Moghaddas, C. Edminster, C. E. Fiedler, S. Haase, M. Harrington, J.

E. Keeley, E. E. Knapp, J. D. McIver, K. Metlen, C. N. Skinner, and A.

142

Youngblood. 2009a. Fire treatment effects on vegetation structure, fuels, and potential fire severity in western US forests. Ecological Applications 19:305-320.

Stephens, S. L., J. J. Moghaddas, B. R. Hartsough, E. E. Y. Moghaddas, and N. E.

Clinton. 2009b. Fuel treatment effects on stand-level carbon pools, treatmentrelated emissions, and fire risk in a Sierra Nevada mixed-conifer forest. Canadian

Journal of Forest Research 39:1538-1547.

Strittholt, J. R., and J. Tutak. 2009. Assessing the Impact of Ecological and

Administrative Considerations on Forest and Shrubland Biomass Projections for

California: NRDC California Biomass Assessment. Conservation Biology

Institute, Corvallis, OR.

Turner, D. P., D. Ritts, B. E. Law, W. Cohen, Z. Yang, T. Hudiburg, J. L. Campbell, and

M. Duane. 2007. Scaling net ecosystem production and net biome production over a heterogeneous region in the western United States. Biogeosciences 4:597-612.

Upton, B., R. Miner, M. Spinney, and L. S. Heath. 2008. The greenhouse gas and energy impacts of using wood instead of alternatives in residential construction in the

United States. Biomass and Bioenergy 32:1-10.

USDA. 2010. Sierra Nevada Forest Plan Amendment Final Supplement Environmental

Impact Statement. USDA Forest Service Pacific Southwest Region, Vallejo, CA

94592.

USGS. 2009. The National Map LANDFIRE: LANDFIRE National Existing Vegetation

Type layer. Pages http://gisdata.usgs.net/website/landfire/ in U.S. Department of

Interior, Geological Survey. U.S. Department of Interior, Geological Survey. van Heerwaarden, L. M., S. Toet, and R. Aerts. 2003. Current measures of nutrient resorption efficiency lead to a substantial underestimation of real resorption efficiency: facts and solutions. Oikos 101:664-669.

Van Tuyl, S., B. E. Law, D. P. Turner, and A. I. Gitelman. 2005. Variability in net primary production and carbon storage in biomass across Oregon forests--an assessment integrating data from forest inventories, intensive sites, and remote sensing. Forest Ecology and Management 209:273-291.

Van Wagner, C. E. 1968. The line-intersect method in furest fuel sampling. Forest

Science 14:20-26.

Waddell, K. L. 2002. Sampling coarse woody debris for multiple attributes in extensive resource inventories. Ecological Indicators 1:139-153.

143

Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. Warming and

Earlier Spring Increase Western U.S. Forest Wildfire Activity. Science 313:940-

943.

Whitaker, J., K. E. Ludley, R. Rowe, G. Taylor, and D. C. Howard. 2010. Sources of variability in greenhouse gas and energy balances for biofuel production: a systematic review. GCB Bioenergy 2:99-112.

Wiedinmyer, C., and M. D. Hurteau. 2010. Prescribed fire as a means of reducing forest carbon emissions in the western United States. Environmental Science &

Technology 44:1926-1932.

Wiedinmyer, C., B. Quayle, C. Geron, A. Belote, D. McKenzie, X. Zhang, S. O'Neill, and K. K. Wynne. 2006. Estimating emissions from fires in North America for air quality modeling. Atmospheric Environment 40:3419-3432.

Williams, R. B. 2007. An Assessment of Biomass Resources in California. California

Biomass Collaborative, California Energy Commission, Davis, CA.

Wirth, C., K. Ogle, A. Nuske, J. T. Weedon, W. K. Cornwell, and J. Cornelissen. 2010.

Global analysis of wood decomposition – the roles of environment, taxonomy, and traits. in . Max Planc Institute for Biogeochemistry.

Wright, L., B. Boundy, R. Perlack, S. C. Davis, and B. Saulsbury. 2006. Biomass energy data book: edition 1, 1 edition. Oak Ridge National Laboratory, Oak Ridge,

Tennessee, USA.

Table A1.

Ecoregion characteristics including dominant forest types, mean annual temperature (MAT), mean annual precipitation (MAP), and area weighted mean fire return interval (MFRI).

144

Ecoregion

1

BM

CB

CO

CP

CR

CV

EC

KM

MB

NB

NC

NR

PL

SB

SM

SN

SR

Forest Dominant Forest Types MAT MAP MFRI

(ha)

3312268 Mixed Conifer, Ponderosa Pine, Juniper, Spruce-Fir 7.3

352650 Pinyon-Juniper, Jeffrey-Ponderosa Pine, Limber-

Bristlecone Pine

6.0

552

445

41 to 45

51 to 60

14.8 652 26 to 30 2688165 Blue Oak-Foothill Pine, Mixed Evergreen, Mixed

Conifer, Redwood, Oak Woodland and Savanna,

Black Oak-Conifer

253667 Mixed Conifer, Ponderosa Pine, Riparian 9.7 330 51 to 60

11.0 1742 91 to

100

4812627 Douglas-fir-Western Hemlock, Sitka Spruce,

Redwood, Mixed Evergreen, Riparian, Western Red

Cedar, Western Hemlock-Silver Fir

170243 Blue Oak-Foothill Pine, Riparian, Salt Desert Scrub,

Mixed Oak Savanna

3545116 Ponderosa Pine, Mixed Conifer, Montane Riparian,

Juniper, Jeffrey Pine-Ponderosa Pine, Lodgepole

Pine, Red Fir, Mountain Hemlock

17.2

9.1

412

630

36 to 40

36 to 40

3748465 Mixed Conifer, Mixed Evergreen, Red Fir, Douglasfir-Western Hemlock, Riparian, Black Oak-Conifer,

Redwood, Mixed Oak Woodland

11.5 1549 16 to 20

93889 Pinyon-Juniper, Montane Riparian, Mixed Oak

Woodland

478106 Juniper, Aspen, Pinyon-Juniper, Montane Riparian,

Jeffrey-Ponderosa Pine, Mountain Mahogany

2311424 Western Hemlock-Silver Fir, Mixed Conifer,

730051 Mixed Evergreen, Mixed Conifer, Mixed Oak, Blue

Oak-Foothill Pine, Oak Woodland, Riparian

18.4

9.7

5.6

Mountain Hemlock, Spruce-Fir, Western Red Cedar,

Riparian, Subalpine Woodland

1514359 Mixed conifer, Riparian, Spruce-Fir, Ponderosa Pine 7.5

1102015 Douglas-fir-Western Hemlock, Riparian, Western

Red Cedar, Sitka Spruce

2175 Pinyon-Juniper

613 51 to 60

10.6 1304 151 to

200

22.0

185

304

1548 101 to

110

41 to 45

51 to 60

125

41 to 45

12.3 1064 26 to 30

8.2 915 21 to 25 1022645 Mixed Conifer, Red Fir, Jeffrey-Ponderosa Pine,

Riparian, Mixed Oak, Subalpine Woodland, Blue

Oak-Foothill Pine, Black Oak-Conifer, Lodgepole

Pine

8613 Montane Riparian 9.7 303 101 to

125

WC 4329871 Douglas-fir-Western Hemlock, Silver Fir-Western

Hemlock, Mountain Hemlock, Mixed Conifer, Red

Fir, Riparian, Western Red Cedar

8.8 1688 101 to

125

WV 538681 Douglas-fir-Western Hemlock, Riparian 11.0 1280 46 to 50

1

BM, Blue Mountains; CB, Central Basin; CO, California Chaparral and Oak Woodlands; CP, Columbia Plateau; CR,

Coast Range; CV, Central California Valley; EC, East Cascades; KM, Klamath Mountains; MB, Mohave Basin; NB, North

Basin and Range; NC, North Cascades; NR, Northern Rockies; PL, Puget Lowlands; SB, Sonoran Basin; SM, Southern

California Mountains; SN, Sierra Nevada; SR, Snake River; WC, West Cascades; WV, Willamette Valley.

Table A2 . State total and mean carbon fluxes for Business-As-Usual (BAU) and treatments. Net biome production (NBP) calculated for 75% treatment efficiency

(possible fire emission reductions). Uncertainty is noted by the ‘±’ symbol.

State

(forested hectares)

Annual Fossil Fuel Emissions

(Tg C yr

-1

)

Carbon density

(Mg C ha

-1

)

Net Primary Production (NPP)

(Tg C yr

(g C m

-1

-2

yr

)

-1

)

Net Ecosystem Production (NEP)

(Tg C yr

(g C m

-1

-2

yr

)

-1

)

Harvest emissions (Tg C yr

-1

)

Fire emissions (Tg C yr

-1

)

Net Biome Production (NBP)

(Tg C yr

(g C 2

-2

-1

yr

)

-1

)

Area Treated (hectares)

Fire Prevention

Economically Feasible

Bioenergy Production

Additional Removals (Tg C yr

-1

)

Fire Prevention

Economically Feasible

Bioenergy Production

Scenario NBP (Tg C yr

-1

)

Fire Prevention

Economically Feasible

Bioenergy Production

Washington

9.0 x 10

6

21

172 ± 25

46.7 ± 4.7

518 ± 52

11.3 ± 1.2

125 ± 13

5.5 ± 0.4

0.9 ± 0.1

4.8 ± 1.3

53 ± 14

0.8 x 10

6

0.8 x 10

7.2 x 10

6

6

0.6 ± 0.02

0.9 ± 0.04

13.2 ± 0.4

4.8 ± 1.3

4.5 ± 1.3

-6.1 ± 1.3

Oregon

12.2 x 10

6

15

150 ± 22

60.0 ± 6.0

488 ± 49

15.2 ± 1.6

125 ± 13

6.4 ± 0.5

1.3 ± 0.2

7.5 ± 1.7

61 ± 14

4.0 x 10

4.0 x 10

9.8 x 10

6

6

6

3.8 ± 0.2

5.7 ± 0.3

17.2 ± 0.7

5.2 ± 1.7

3.3 ± 1.7

-6.6 ± 1.9

California

12.8 x 10

6

105

130 ± 18

61.0 ± 6.2

477 ± 48

18.1 ± 2.1

142 ± 16

2.7 ± 0.2

1.8 ± 0.3

13.6 ± 2.1

107 ± 16

6.8 x 10

6.8 x 10

7.9 x 10

6

6

6

6.7 ± 0.4

10.5 ± 0.7

13.4 ± 0.9

9.4 ± 2.1

5.6 ± 2.2

2.9 ± 2.3

145

Table A3 . Forest types that contribute more than 1 g C m

-2

y

-1

to the observed area-weighted difference of 46.4 g C m

-2

y

-1

(see also Table 2) in net biome production (NBP) between California and Oregon for shared forest types. For example, this difference in NBP is partially due to an area-weighted difference in net ecosystem production (NEP) and harvest of 19.2 and

2.7 g C m

-2

y

-1 for the California Coastal Redwood forest type (see example calculation in footnote). Units are in g C m

-2

y

-1 unless otherwise noted.

Forest Type

North Pacific Maritime Mesic-Wet

Douglas-fir-Western Hemlock Forest

State NEP Fire Harvest

CA

OR

189

205

4

0

0

Area

(ha)

3488

113 2684324

ΔNEP

-44.9

ΔFire ΔHarvest

-0.1 -24.9

Weighted

NBP

0.1

19.9

ΔNBP

-19.9

CA 296 1 41 832912 16.6

California Coastal Redwood Forest 19.2 0.1 2.7 16.5

California Lower Montane Blue Oak-

Foothill Pine Woodland and Savanna

Mediterranean California Mesic

Mixed Conifer Forest and Woodland

Central and Southern California

Mixed Evergreen Woodland

California Montane Riparian Systems

OR

CA

OR

CA

OR

CA

OR

CA

238

120

0

111

107

186

0

165

0

13

14

28

0

14

0

7

26 2983

17 1636673

0 0

27 2748963

29 721488

2 643670

0 0

0 667775

15.4

17.6

9.4

8.1

0.9

1.9

1.4

0.7

2.2

4.1

0.1

0.0

0.1

12.3

0.0

15.4

3.8

7.9

0.0

7.9

12.3

11.6

7.9

7.4

Mediterranean California Dry-Mesic

Mixed Conifer Forest and Woodland

Mediterranean California Mixed

Evergreen Forest

North Pacific Maritime Dry-Mesic

Douglas-fir-Western Hemlock Forest

Great Basin Pinyon-Juniper

Woodland

Mediterranean California Mixed Oak

Woodland

OR

CA

OR

CA

OR

CA

OR

CA

OR

CA

OR

163

153

158

308

228

166

163

208

221

155

151

18

20

8

11

76

0

11

23

16

27

7

0 42589

27 1559888

37 671005

62 581798

38 562982

0 3184

78 838685

0 343308

0 8570

0 455384

0 9823

10.1

3.6

-11.1

5.4

5.4

2.0

-3.0

-0.7

0.6

1.0

1.2

1.1

-5.4

0.0

0.0

0.5

13.0

6.2

10.7

5.2

0.0

5.1

5.0

0.1

4.5

0.1

6.8

5.5

-5.0

4.8

4.4

Forest Type

Mediterranean California Red Fir

Forest

North Pacific Hypermaritime

Sitka Spruce Forest

Northern Rocky Mountain

Ponderosa Pine Woodland and

Savanna

North Pacific Montane Riparian

Woodland and Shrubland

Mediterranean California Lower

Montane Black Oak-Conifer

Forest

Southern California Oak

Woodland and Savanna

Northern Rocky Mountain Dry-

Mesic Montane Mixed Conifer

Forest

California Montane Jeffrey Pine(-

Ponderosa Pine) Woodland

State

CA

OR

CA

OR

CA

OR

CA

OR

CA

OR

CA

OR

CA

OR

CA

OR

NEP Fire Harvest Area (ha) ΔNEP

91

90

0

275

59

63

13

2

0

0

3

10

20

38

0

145

21

20

1087841

147570

0

321231

104182

1225805

6.7

-7.2

-5.8

178

175

158

159

158

0

41

40

71

51

0

4

7

13

9

3

3

11

0

0

20

17

0

0

27

23

22

2846

192321

320462

28108

261241

0

4468

1737309

601572

71 89 22 13831

-2.7

3.6

3.2

-5.6

3.3

ΔFire ΔHarvest

1.1

0.0

-0.9

-0.1

0.3

1.0

-1.0

0.4

1.3

-3.8

-1.9

0.0

0.6

0.0

-2.4

1.0

Weighted

NBP

4.9

0.6

0.0

3.4

0.3

3.3

2.2

0.0

0.0

2.2

0.0

2.7

2.9

0.3

1.8

0.0

ΔNBP

4.3

-3.4

-3.0

-2.7

2.6

2.2

-2.2

1.9

North Pacific Dry-Mesic

Silver Fir-Western

Hemlock-Douglas-fir

Forest

CA

OR

122

142

0

7

0

73

9

350817

-4.1 -0.2 -2.1

0.0

1.8

-1.8

North Pacific Swamp Systems

CA

OR

190

190

0

5

0

0

71

112087

-1.7 0.0 0.0

0.0

1.7

-1.7

North Pacific Mountain Hemlock

Forest

CA

OR

105

99

2

14

33

48

9232

372207

-2.9 -0.4 -1.4

0.1

1.1

-1.1

1

Weighted NBP = ((NEP – Fire – Harvest) * Forest Type Area) / Total Forest Area of State; Total forest areas are 12.2 and 12.8 million hectares for Oregon and California, respectively (Table 2)

Table A4.

Life Cycle Assessment coefficients. Each coefficient is multiplied by a biomass pool (Tg C yr

-1

). Processing efficiencies vary by state (PNW = Pacific Northwest or Oregon and Washington; CA = California)

Life-Cycle Factor Biomass Pool Description and source

Sawlog (SL)

Coefficien t

0.80 Merchantable wood Sawlog fraction of merchantable portion of harvest(Smith et al. 2006)

Pulpwood (PW)

WD1a

0.11

0.09

Merchantable wood

Merchantable Wood

Sawlog (SL)

Pulpwood fraction of merchantable portion of harvest(Smith et al. 2006)

Portion of wood lost in initial manufacturing process(Smith et al. 2006)

Wood product fraction of sawlog (Smith et al. 2006) Sawlog Wood (PNW)

Sawlog Wood (CA)

Sawlog Paper (PNW)

Sawlog Paper (CA)

Pulpwood Paper

(PNW)

Pulpwood Paper (CA)

WD1b

0.56

0.53

0.25

0.145

0.50

0.145 na

Sawlog (SL)

Pulpwood (PW)

(SL+PW)- (Wood + Paper)

Paper product fraction of sawlog (Smith et al. 2006)

Paper product fraction of pulpwood (Smith et al.

2006)

WD2

Wood Industry FFE

Harvest

Manufacturing

FF WTT

3

2

Harvest Transport

Wood

Paper

Product Transport

0.46

.009

.003

0.004

0.57

0.009

0.1675

(Wood + Paper)

Merchantable Wood

Merchantable Wood

Wood

Paper

(Wood + Paper)

Wood Industry FFE

Portion of wood lost in conversion to products

(Smith et al. 2006)

10yr Decomposition (no longer in use or in a landfill) (Smith et al. 2006)

Harvest FFE (Sonne 2006)

Transport FFE (Heath et al. 2010) for average 75 km distance(Evans and Finkral 2009)

Wood FFE (Heath et al. 2010)

Paper FFE (Heath et al. 2010)

Transport FFE (Heath et al. 2010) for average 250 km distance(Evans and Finkral 2009)

Fossil Fuel extraction and production (ICCT 2010)

Bioenergy FFE

Combustion (CHP)

Cellulosic Ethanol

(CE)

Bioenergy Substitution

0.05

0.35

Non-Merchantable Wood

Non-Merchantable Wood

Non-Merchantable Wood

Non-Merchantable Wood

Harvest, Transport, Chip manufacturing(Heath et al.

2010)

Harvest, Transport, Conversion to Ethanol(Jaeger et al. 2007, Whitaker et al. 2010)

FFE reduction for the energy potential of wood energy compared to fossil fuel (Mitchell et al. 2009) Cellulosic Ethanol

(CE)

0.36

FF WTT Displacement 0.1675 Bioenergy Substitution Displaced FF emissions from crude oil extraction

Bioenergy Emissions 1.00

Wood Substitution 0.36

Non-Merchantable Wood

Sawlog Wood

C released from bioenergy use

FFE reduction with 50:50 mix of aluminum/steel substitution(Upton et al. 2008)

1

LCA = FCO

2

= NBP + Total Harvest – WD1a – WD1b – WD2 – Wood Industry FFE - Bioenergy Emissions + Bioenergy Substitution +

FF WTT displacement + (Wood Substitution); WTT = Well to tank emissions; WD = Wood decomposition; Wood Substitution is

2 dependent on increase in wood use

3

FFE = Fossil Fuel Emissions

WTT = Well to Tank emissions from crude oil extraction are approximately 15% of total well to wheels (extraction plus use) emissions.

Fire emissions

Total harvest

3

Current

Additional

NBP

4

Wood

Product

FFE (fossil fuel emissions)

Wood

Decompositi on (WD1)

Wood

Decompositi on

(WD2)

Bioenergy emissions

FF

Substitution

FF WTT

Displaced

5

Wood

Substitution

FCO

2

6

No Wood

Sub

With Wood

Sub

Uncertainty

Σ Tg C added to atmosphere

(20 yrs)

No Wood

Sub

With Wood

Sub

150

Table A5.

Life cycle assessment of forest derived bioenergy for the West Coast region assuming a 50:50 mix of combustion for combined heat and power (CHP) and conversion to cellulosic ethanol (CE). All values are in Tg C yr

-1

unless otherwise noted. Numbers in bold represent optimum conditions, biomass pool specific combustion factors, and no wood substitution.

LCA

Parameter

NEP

Business-as-

Usual (BAU)

Opt | Alt | Low

1

44.5

Fire

Prevention

Opt | Alt | Low

38.0

Economically

Feasible

Opt | Alt | Low

38.0

Bioenergy

Production

Opt | Alt | Low

30.5

Regrowth

2

0.00 9.6 9.6 20.6

4.1 | 8.1 | na 2.5 | 3.9 | 3.0 2.5 | 3.9 | 3.0 2.5 | 3.9 | 3.0

14.6

0.00

25.8 | 21.8 | na

1.06

4.7

4.5

0.00

0.00

0.00

0.00

28.9 | 24.8 | na na

3.0

0.00

0.00

14.6

11.1

19.4 | 18.0 | 18.9

2.74

9.34

4.25 | 4.25 | 4.68

6.80

3.9 | 3.9 | 2.5

0.65 | 0.65 | 0.42

0.7

26.6 | 25.1 | 24.0

27.2

3.1

46.0 | -7.0 | 97.3

32.9

14.6

17.1

13.4 | 12.0 | 12.9

6.69

10.46

4.64 | 4.64 | 4.96

10.65

6.1 | 6.1 | 3.9

1.01 | 1.01 | 0.67

1.0

19.8 | 18.4 | 16.1

20.8

3.3

174.6 | 127.7 | 254.4

161.1

14.6

43.8

-9.7 | -11.1 | -10.3

8.07

16.82

7.12 | 7.12 | 9.16

24.12

13.7 | 13.7 | 8.9

2.30 | 2.30 | 1.49

3.1

8.6 | 7.2 | 0.4

11.6

3.5

420.7 | 352.0 | 569.1

344.5

Table Footnotes (continued)

1

Opt = Optimum efficiency parameters, Alt = Alternative combustion coefficients, not

151 pool specific, Low = Low efficiency parameters

2

Thinned plot total NEP is the sum of the NEP and Regrowth rows

3

Total harvest includes the continued harvest for wood products plus an additional harvest for bioenergy resulting in harvest increases compared to BAU for each scenario

4

NBP = Scenario NEP + Scenario Regrowth – Scenario Total harvest – Scenario Fire emissions

5

FF WTT = Fossil fuel well-to-tank emissions from extraction and refining of crude oil

6

FCO

2

= NEP + Regrowth – Fire –Wood Product FFE (includes associated FF WTT) –

WD1 – WD2 – Bioenergy Emissions + FF Substitution + FF WTT displacement +

(Wood Substitution)

Table A6. Sensitivity Analysis parameters and relative impact on results (CHP = biomass used as combined heat and power; FP = Fire Prevention scenario; EC =

Economically Feasible scenario; BP = Bioenergy Production scenario).

Efficiency

Optimum

Moderate

Low

Wood

Substitution

Parameters evaluated Value

CHP substitution 0.78 efficiency

Ethanol substitution 0.36 efficiency

Percentage wood products still in use or in landfill

Fire emissions reductions

0.64

75%

CHP substitution efficiency

Ethanol substitution efficiency

Percentage wood products still in use or in landfill

Fire emissions reductions

0.68

0.26

0.54

75%

CHP substitution efficiency

Ethanol substitution efficiency

Percentage wood products still in use or in landfill

Fire emissions reductions

0.58

0.16

0.44

50%

Combustion Woody biomass combustion coefficients

Included wood substitution in LCA

Impact on result (%)

FP +3%

EC +5%

BP +13%

These are the optimum or best case scenario conditions and were used for the results. The other sets of parameter values are compared to the FCO

2

value obtained for this parameter set.

Negative values indicate emissions savings.

FP +9%

EC +12%

BP +28%

30% combustion of all aboveground woody biomass

36% substitution benefit of structural wood products pool

FP -9%

EC -6%

BP +1%

FP -2%

EC -3%

BF -10%

152

Table A7. Forest Inventory Analysis plot information for the study region.

Ecoregion

BM

CB

CO

CP

CR

CV

EC

KM

MB

1

Number of

Plots

885

129

672

75

1057

13

1097

1154

57

With

Increment

Data

857

126

633

66

1036

11

1069

1137

53

With Phase 3 measurements

672

101

509

48

819

10

855

878

42

Percentage excluded

22

20

20

27

21

9

20

23

21

NB

NC

NR

PL

115

482

338

174

106

477

334

169

76

359

258

122

28

25

23

28

SM

SN

204

1275

192

1258

167

1035

15

18

WC 1019 998 801 20

WV 115 109 91 17

1

BM, Blue Mountains; CB, Central Basin; CO, California Chaparral and Oak Woodlands; CP,

Columbia Plateau; CR, Coast Range; CV, Central California Valley; EC, East Cascades; KM,

Klamath Mountains; MB, Mohave Basin; NB, North Basin and Range; NC, North Cascades; NR,

Northern Rockies; PL, Puget Lowlands; SB, Sonoran Basin; SM, Southern California Mountains;

SN, Sierra Nevada; SR, Snake River; WC, West Cascades; WV, Willamette Valley.

153

154

Table A8 . Equations and factors used to calculate carbon stocks, fluxes, and life-cycle assessment.

Component

Bole Biomass

Bark biomass biomass

Volume

3

1

Branch biomass

Foliage biomass

Coarse root biomass

Leaf Area Index

2

Fine root biomass

Understory shrub

Coarse woody debris

Coarse woody debris biomass

Code

Biomass

Biomass

Biomass

Biomass

Biomass

LAI

Biomass

Biomass

Volume

Biomass b br ba f cr fr s cwd cwd

Equation

Bole Volume

Equation*Wood Density

Allometric Equation

Allometric Equation

Allometric Equation

Allometric Equation

Biomass

(exp(4.4179+(.3256*LAI

)-(.0237*LAI

2

)) a*(1-(exp(-b*Shrub

Volume)))

(9.869/(8*L))*(D

Volume f

/ LMA cwd

Density

2

)

5

* Adjusted

Notes

Bole volume derived from allometric equation using

DBH (diameter at breast height) and height; (Jenkins et al. 2003, Hudiburg et al.

2009)

Derived from DBH and/or height;(Jenkins et al. 2003,

Hudiburg et al. 2009)

Derived from DBH and/or height; (Jenkins et al. 2003,

Hudiburg et al. 2009)

Derived (allometry) from

DBH and/or height;

(Jenkins et al. 2003,

Hudiburg et al. 2009)

Derived from a volume equation developed for

Douglas-fir and speciesspecific wood densities;

(Jenkins et al. 2003,

Hudiburg et al. 2009)

Calculated from Biomass f and leaf mass per unit leaf area (LMA)

(Van Tuyl et al. 2005),

Supplemental plot data

(p<0.001, R

2

= 0.41, n=36)

Shrub volume calculated as the product of the recorded fraction plot cover, plot area, and height

Where, L is the transect length in meters and D is the diameter of the piece in centimeters (Van Wagner

1968, Harmon and Sexton

1996, Waddell 2002)

Derived by multiplying

Volume cwd

by a decay class adjusted species-specific density

155

Component

Wood NPP (Bole,

Branch, Bark, and

Coarse Roots)

Foliage NPP

Fine root NPP

NEP

4

Code

NPP

NPP

NPP

NEP w f fr

Equation Notes

Biomass w2

– Biomass w1

Difference between biomass of woody components at current and previous time steps

Biomass f

/ Foliage

Retention time

Biomass of foliage divided by the average number of years of foliage a stand carries

5

Biomass fr

* 1.2 year

-1

Average fine root turnover

(1.2 year-1) obtained from the literature and supplemental plot data

ANPP – Rh wood

– litterfall

+ Δ root + Δ soil C

(Keyes and Grier 1981,

Campbell et al. 2004a)

Where Δ fine root and Δ soil

C are assumed to be zero over the time period(Giardina and Ryan

2002)

Litterfall

Dead Wood Rh Rh wood

NPP f

x 0.79

Biomass cwd

*exp(-kt)

- Biomass cwd

Average mass retention of

21%; (van Heerwaarden et al. 2003)

Where k value is calculated as a function of piece size, genus, precipitation, and mean annual temperature ( http://www.bgcjena.mpg.de/bgcorganisms/pmwiki.php/Rese arch/FET)

1

Species-specific wood densities were obtained from US Forest Service wood density survey for western Oregon (Maeglin and Wahlgren 1972), the Forest Products Laboratory wood handbook

(1974) (Laboratory 1974), and from wood cores obtained on our supplemental plots. Wood densities were reduced according to decay class for standing dead trees (Waddell 2002).

2

Leaf specific mass (LMA) was obtained from a look-up Table of species-specific values obtained from measurements on the supplemental plots in each of the ecoregions. In some cases, a species-specific value was not available and therefore a closely related species was used.

3

The parameters ‘a’ and ‘b’ are regression coefficients that vary by species. Equations were developed from harvested shrubs at the supplemental field plots.

4

Foliage (branch) samples from evergreen species were collected at all supplemental plots and the average number of years of growth retained on each branch was recorded to calculate retention time. Samples were also dried and weighed. New shoot growth was recorded for foliage NPP.

This information was used to construct species and ecoregion specific lookup Tables for the FIA plots.

Table A9.

Comparison of combustion factors by source and fuel category.

Source Forest

Types

Fuel

Category

Combustion Factor (Fraction

Combusted) by severity

High Moderate Low Unburned/Very

Campbell et al., 2007

Mixed

Conifer,

Douglas-

Trees

Snags

0.08

0.18

0.07

0.14

Low

0.03 0.00

0.11 0.01 fir,

Western

Hemlock,

Tanoak,

Shrubs

Foliage

FWD

CWD

0.86

1.00

1.00

0.78

0.66

0.76

0.76

0.58

0.42 0.00

0.75 0.70

0.75 0.70

0.61 0.62

Meigs et al.,

2009

1

Wiedinmyer

Jeffrey

Pine

Ponderosa

Pine n.a.

Litter

Duff

Live Trees

1.00

0.99

0.03

0.76

0.51

0.014

0.75 0.70

0.54 0.44

.003 n.a. and Hurteau

(2010)

2

,

Wiedinmyer

Aboveground woody mass

Litter/Duff

0.30

0.90

0.30

0.90

0.30 0.30

0.90 0.90 et al., 2006

1

From Consume 3.0 simulations (Ottmar et al. 2006) and field measurements of consumption

2

Combustion factors were not indicated to vary by severity in the reported citations

156

157

Figure A1 . A . Thinning treatment scenarios. Included are the percent basal area reductions, the maximum tree diameter (DBH) that can be harvested, and the land area where the proposed treatment would be implemented. The treatments remove enough live biomass in order to lower risk of wildfire and provide biomass for bioenergy. MFRI is the mean fire return interval. B . Boundary of the processes accounted for in the life-cycle assessment (LCA). The boundary was expanded to account for substitution of fossil fuels by bio-energy. Full lines show C flow and dotted lines show energy flow. Arrows show fluxes and lines show substitution. Carbon is exchanged between the forest and the atmosphere through photosynthesis (1) and respiration (2) and lost to the atmosphere via fire (2) or removed by harvest (3). The carbon removed is used for bioenergy or wood products. Transport of the biomass to either end use utilizes fossil fuels and contributes to the total fossil fuel emissions (FFE) (4). FFE are associated with both manufacturing of wood products and both forms of bioenergy production (energy is required to convert the biomass to a useable form of energy; (5)). Biomass utilized for wood products can end up in a long term storage product (structural wood), a short term product (paper), imported, or exported. Some wood product carbon reenters the atmosphere through slow (6; wood) or rapid (7; paper) decomposition or combustion while some of it is eventually disposed in landfills (8 and 9) where it is very slowly decomposed. Biomass utilized for bioenergy can be burned or converted to cellulosic ethanol, releasing carbon to the atmosphere (10).

Wood products can be substituted for fossil fuel products (11) and bioenergy emissions can be substituted for fossil fuel emissions associated with use, extraction, and production

(12).

158

Figure A2. (A) MFRI (Mean Fire Return Interval) verses delta CO

2

(Fire Prevention; FP

– Business as Usual; BAU) and (B) initial emissions versus delta CO

2

.

159

Figure A3 . Conceptual analysis of harvest to NEP ratio, product mix, and varying efficiencies of bioenergy production. The x-axis is the harvest to NEP ratio, the y-axis is the product to harvest ratio (a value of 1 indicates all harvest is in wood products, a 0 indicates all harvest is used for bioenergy). F

0

lines (black and grey) are the combinations for carbon-neutral flux and F

BAU

(dark red and red) are the lines where the flux is equal to

BAU. A) Energy mix is 50:50. B) Energy mix is varied. Dark red line = BAU and red line

= where the energy mix ratio flux is equal to BAU. Black line = 100% biomass combustion and the grey line = 100% cellulosic ethanol conversion. Nearly all of the harvest must be used for bioenergy to realize a lower emission than BAU. C) Fossil fuel inputs are varied. Black line = fossil fuel inputs are equivalent to product biomass (least efficient) and grey line = No fossil fuel inputs required (most efficient). Even if no fossil fuel inputs are required, only a small increase in harvest decreases emissions compared to

BAU. D) Amount of wood in short term pool is varied and used for bioenergy instead of going to landfill. Black line = 100% of wood product enters short term pool and then used for bioenergy and the grey line = 100% of wood product enters long term pool. With complete recycling of wood waste (no short term pool) for use as bioenergy, there is still very little emissions savings.

Appendix B: Supporting information for Chapters 3 and 4

Methods for formatting regional downscaled climate forcing dataset for use by CLM4

The regional downscaled dataset includes daily precipitation, minimum and maximum

160 temperature, and wind speed. To format the datasets for use by CLM4, the required shortwave radiation and relative humidity were calculated incorporating algorithms from

DAYMET (Thornton et al. 1997), and methods for sub-daily calculations as described by

(Göeckede et al. 2010). We created hourly atmospheric forcing data files to be used in offline CLM4 simulations. Source data files are from the climate impacts group (Salathe

2008) and are downscaled historical and future ECHAM5 AR2 files specifically designed for the Pacific Northwest http://www.cses.washington.edu/data/ipccar4/ .

The specific variable calculations are as follows:

1.

Air temperature: The regional input files provide daily minimum and maximum temperature. Daylength was used to apply a diurnal pattern to the minimum and maximum temperatures.

2.

Wind speed: The regional input files provide daily estimates of wind in m/s. Wind speed was assumed to be constant for sub-daily time steps.

3.

Relative humidity: Percent relative humidity was calculated using the hourly mean temperatures calculated above and vapor pressure. a.

DAYMET and MTClim algorithms were used to calculate vapor pressure from temperature, precipitation, and solar radiation. Vapor pressure was then used to calculate relative humidity (RH): i.

RH = 100 * (VP/SVP); where VP = the average daily vapor pressure in Pascals and SVP = the saturation vapor pressure. SVP varies with temperature. ii.

SVP = 610.78 * exp(T/(T+238.3)*17.2694); where T is the current temperature in degrees C.

4.

Precipitation: The input files provide daily sums of precipitation. This needed to be distributed over the day, but not evenly. CLM will evaporate off the water too quickly and none of it will reach the plant roots. The precipitation was split into 3 equal

161 amounts of precipitation and dropped at 8 hour intervals similar to the NCEP dataset where it is dropped at 6 hour intervals. We recognize more sophisticated diurnal precipitation algorithms using site observations could be developed, but more locations with sub-daily patterns of rainfall would be necessary for the region.

6.

FSDS: Incoming shortwave radiation or incident solar is not provided in the input files. Again, DAYMET algorithms were used. The inputs required are daily Tmin,

Tmax, precipitation, latitude, longitude, and elevation all of which are available from the downscaled regional dataset.

Methods for making PFT modifications in CLM4

There are 17 PFTs in CLM4, eight of which are forested. Because we did not need the tropical or boreal PFTs, we reorganized the PFT physiological file to represent the variation between and within the temperate evergreen needleleaf PFT among the ecoregions. The old and new PFT assignments are shown in Table B1 and Table B2 below. We also show the parameters that varied for each new PFT. The surface datasets and the dynamic PFT file were altered to match the new PFT assignments. Inventory and remote sensing data were used to identify within region PFT percentage cover for ‘fir’ versus ‘pine’ where necessary.

Methods for stem allocation modification in CLM4

The PFT physiology file includes an option to have the fraction of annual NPP allocated to stem wood versus foliage change with the increasing annual sum of NPP throughout the year. The equation is for all forested PFTs:

Allocation ratio = (2.7 / (1.0+exp (-0.004*(annsum_npp(t) - 300.0)))) - 0.4 Eq. 1

Where ‘annsum_npp’ is the total PFT NPP summed over the year at the current timestep(t). The ratio is constrained to be 0.20 when NPP = 0 and does not exceed 2.2 for

NPP values greater than 1000 g C m

-2

yr

-1

. Using inventory data, we found annual stem wood to foliage allocation ratios as high as 3.0 for the mesic ecoregions and no higher than 2.0 for the semi-arid ecoregions. We were unable to investigate seasonal allocation

ratios with the inventory data. We modified the equation to increase allocation to stem

162 wood for lower values of NPP in the Coast Range, West Cascades and Klamath

Mountains (mesic) and decrease allocation to stem wood for all values of NPP in the

Blue Mountains and East Cascades (semi-arid; Figure B1 and Table B3).

References

Eidenshink, J. et al. A project for monitoring trends in burn severity. Fire ecology 3, 3-20

(2007).

French, N. H. F., W. J. de Groot, L. K. Jenkins, B. M. Rogers, E. Alvarado, B. Amiro, B. de Jong, S. Goetz, E. Hoy, E. Hyer, R. Keane, B. E. Law, D. McKenzie, S. G.

McNulty, R. Ottmar, D. R. Pérez-Salicrup, J. Randerson, K. M. Robertson, and

M. Turetsky. 2011. Model comparisons for estimating carbon emissions from

North American wildland fire. J. Geophys. Res. 116:G00K05.

Göeckede, M., A. M. Michalak, D. Vickers, D. P. Turner, and B. E. Law. 2010.

Atmospheric inverse modeling to constrain regional-scale CO2 budgets at high spatial and temporal resolution. J. Geophys. Res. 115:D15113.

Salathe, E. P. 2008. Daily downscaled data for the Pacific Northwest. in . Climate Impacts

Group, University of Washington.

Thornton, P. E., S. W. Running, and M. A. White. 1997. Generating surfaces of daily meteorological variables over large regions of complex terrain. J. Hydrol.

190:214–251. van der Werf, G. R., J. T. Randerson, L. Giglio, G. J. Collatz, M. Mu, P. S. Kasibhatla, D.

C. Morton, R. S. DeFries, Y. Jin, and T. T. van Leeuwen. 2010. Global fire emissions and the contribution of deforestation, savanna, forest, agricultural, and peat fires (1997–2009). Atmos. Chem. Phys. 10:11707-11735.

Table B1. CLM PFT original configuration and the new subgroup that replaced the original PFT for regional variation in CLM_eco and CLM_stem.

CLM PFT Description

1 Evergreen needleleaf (temperate)

2 Evergreen needleleaf (boreal)

New PFT

CR all, KM fir*

WC fir

3

4

5

6

Deciduous needleleaf (boreal)

Evergreen broadleaf (tropical)

Evergreen broadleaf (temperate)

Deciduous broadleaf (tropical)

7

8

9

10

Deciduous broadleaf (temperate)

Deciduous broadleaf (boreal)

Evergreen broadleaf shrub

Deciduous broadleaf shrub (temperate)

* Fir includes Douglas-fir and true firs

BM all

WC, KM pines

No change

EC fir

No change

EC pines

No change

No change

163

164

Table B2. Model physiological parameters that vary for subgroups of PFTs in each ecoregion

New PFT

(PFT number)

Specific Leaf Area

(SLA; m

-2

/ g C)

Leaf

Longevity

Mortality rate (%)

CR all, KM fir (1) .013

WC fir (2)

BM all (3)

.013

.007

WC, KM pine (4) .009

EC fir (6) .010

EC pine (8) .007

35

48

55

50

43

56

Leaf N

in Rubisco

5.0

5.0

3.5

3.7

3.7

3.7

3

5

3

5

5

3

0.8

1.0

1.3

1.0

0.8

1.0

165

Table B3. Stem wood allocation equations for each ecoregion PFT. The equation was only varied for ecoregions where stem wood was significantly over- or underestimated.

Maximum allocation is the upper limit of annual NPP that will be allocated to stem wood production versus foliage production.

New PFT

(PFT number)

Allocation

Equation

Maximum allocation

CR all, KM fir (1) 2.6/ (1.0+exp(-0.010*(annsum_npp(t) - 100.0))) - 0.25 2.4

WC fir (2)

BM all (3)

2.6 / (1.0+exp(-0.010*(annsum_npp(t) - 100.0))) - 0.25 2.4

2.0 / (1.0+exp(-0.0035*(annsum_npp(t) - 300.0))) - 0.27 1.7

WC, KM pine (4) 2.7 / (1.0+exp(-0.004*(annsum_npp(t) - 300.0))) - 0.4 2.2

EC fir (6) 2.6 / (1.0+exp(-0.010*(annsum_npp(t) - 100.0))) - 0.25 2.4

EC pine (8) 2.7 / (1.0+exp(-0.004*(annsum_npp(t) - 300.0))) - 0.4 2.2

Figure B1. Stemwood allocation (ratio of stem wood allocation to foliage allocation) versus NPP. The default CLM4 equation is blue and the modified equations are for mesic

(Green) and semi-arid (Red).

166

A

167

Figure B2. GPP versus monthly precipitation at the flux towers sites. Red squares are tower observations of GPP and blue diamonds are the CLM simulated monthly GPP values. CLM4 is underestimating GPP for the low values of precipitation (summer months).

168

Figure B3. Historical annual area burned in Oregon (km

-2

). Estimates are from the

Monitoring Trends in Burn Severity database (MTBS; Eidenshink, J. et al. 2007), the

Global Fire Emissions Database (GFED; van der Werf, G. et al. 2010), and CLM4.

CLM4 overestimates burn area in Oregon for all years except for 2002 compared to the remote sensing based estimates. CLM4 does not include a fire suppression algorithm that could be contributing to the high bias. However, burn area estimates do not exceed more than 2% of the land area for any year. GFED estimates are known to be the lowest for burn area compared to other models (French et al. 2011)

Figure B4. Burn area is predicted to double over the next century.

169

170

Figure B5. Predicted annual relative humidity (%) using the downscaled regional climate data and DAYMET algorithms. Relative humidity remains constant.

171

Figure B6. Monthly GPP and monthly precipitation for a period of drier years (2076-

2078) followed by a period of wetter years (2079-2081). For the East Cascades and Blue

Mountains there is a slight increase in summer GPP in the last two years.

172

Figure B7. Metolius tower site observation of annual NPP (from biometric methods) and annual GPP (from eddy-covariance data) versus CLM4 modeled annual NPP and GPP.

Uncertainty in observations is indicated by the black error bars. NPP is underestimated by an average of 24% while GPP is underestimated by and average of 30%.

Download